tanu/LSHTM_analysis
1
0
Fork 0
Code Issues Pull requests Releases Wiki Activity

added ml_functions dir

Browse source
This commit is contained in:
Tanushree Tunstall 2022-06-29 12:06:47 +01:00
parent c85c965c3e
commit 9aadb0329f
30 changed files with 683 additions and 606160 deletions
Download patch file Download diff file Expand all files Collapse all files

73
scripts/ml/combined_model/ml_data_combined
View file

@ -1,73 +0,0 @@
#!/usr/bin/env python3
# -*- coding: utf-8 -*-
"""
Created on Sat Jun 25 11:07:30 2022
@author: tanu
"""
import sys, os
import pandas as pd
import numpy as np
import re
###############################################################################
homedir = os.path.expanduser("~")
sys.path.append(homedir + '/git/LSHTM_analysis/scripts/ml/functions')
###############################################################################
#====================
# Import ML functions
#====================
#from MultClfs import *
from GetMLData import *
from SplitTTS import *
#%% Load all gene files #######################################################
# param dict
combined_model_paramD = {'data_combined_model' : True
, 'use_or' : False
, 'omit_all_genomic_features': False
, 'write_maskfile' : False
, 'write_outfile' : False }
pnca_df = getmldata('pncA', 'pyrazinamide' , **combined_model_paramD)
embb_df = getmldata('embB', 'ethambutol' , **combined_model_paramD)
katg_df = getmldata('katG', 'isoniazid' , **combined_model_paramD)
rpob_df = getmldata('rpoB', 'rifampicin' , **combined_model_paramD)
gid_df = getmldata('gid' , 'streptomycin' , **combined_model_paramD)
alr_df = getmldata('alr' , 'cycloserine' , **combined_model_paramD)
# quick check
foo = pd.concat([alr_df, pnca_df])
check1 = foo.filter(regex= '.*_affinity|gene_name|ligand_distance', axis = 1)
# So, pd.concat will join correctly but introduce NAs.
# TODO: discuss whether to make these 0 and use it or just omit
# For now I am omitting these i.e combining only on common columns
expected_nrows = len(pnca_df) + len(embb_df) + len(katg_df) + len(rpob_df) + len(gid_df) + len(alr_df)
# finding common columns
dfs_combine = [pnca_df, embb_df, katg_df, rpob_df, gid_df, alr_df]
common_cols = list(set.intersection(*(set(df.columns) for df in dfs_combine)))
expected_ncols = np.min([len(pnca_df.columns)] + [len(embb_df.columns)] + [len(katg_df.columns)] + [len(rpob_df.columns)] + [len(gid_df.columns)] + [len(alr_df.columns)])
expected_ncols
if len(common_cols) == expected_ncols:
print('\nProceeding to combine based on common cols (n):', len(common_cols))
combined_df = pd.concat([df[common_cols] for df in dfs_combine], ignore_index = False)
print('\nSuccessfully combined dfs:'
, '\nNo. of dfs combined:', len(dfs_combine)
, '\nDim of combined df:', combined_df.shape)
else:
print('\nFAIL: could not combine dfs, length mismatch'
, '\nExpected ncols:', expected_ncols
, '\nGot:', len(common_cols))
#%% split data into different data types
tts_7030_paramD = {'data_type' : 'actual'
, 'split_type' : '70_30'
, 'oversampling' : True}
data_CM_7030D = split_tts(ml_input_data = combined_df
, **tts_7030_paramD
, dst_colname = 'dst'
, target_colname = 'dst_mode'
, include_gene_name = False) # when not doing leave one group out

833
scripts/ml/functions/MultClfs.py
View file

@ -1,833 +0,0 @@
#!/usr/bin/env python3
# -*- coding: utf-8 -*-
"""
Created on Fri Mar 4 15:25:33 2022
@author: tanu
"""
#%%
import os, sys
import pandas as pd
import numpy as np
import pprint as pp
from copy import deepcopy
from sklearn import linear_model
from sklearn import datasets
from collections import Counter
from sklearn.linear_model import LogisticRegression, LogisticRegressionCV
from sklearn.linear_model import RidgeClassifier, RidgeClassifierCV, SGDClassifier, PassiveAggressiveClassifier
from sklearn.naive_bayes import BernoulliNB
from sklearn.neighbors import KNeighborsClassifier
from sklearn.svm import SVC
from sklearn.tree import DecisionTreeClassifier, ExtraTreeClassifier
from sklearn.ensemble import RandomForestClassifier, ExtraTreesClassifier, AdaBoostClassifier, GradientBoostingClassifier, BaggingClassifier
from sklearn.naive_bayes import GaussianNB
from sklearn.gaussian_process import GaussianProcessClassifier, kernels
from sklearn.gaussian_process.kernels import RBF, DotProduct, Matern, RationalQuadratic, WhiteKernel
from sklearn.discriminant_analysis import LinearDiscriminantAnalysis, QuadraticDiscriminantAnalysis
from sklearn.neural_network import MLPClassifier
from sklearn.svm import SVC
from xgboost import XGBClassifier
from sklearn.naive_bayes import MultinomialNB
from sklearn.preprocessing import StandardScaler, MinMaxScaler, OneHotEncoder
from sklearn.compose import ColumnTransformer
from sklearn.compose import make_column_transformer
from sklearn.metrics import make_scorer, confusion_matrix, accuracy_score, balanced_accuracy_score, precision_score, average_precision_score, recall_score
from sklearn.metrics import roc_auc_score, roc_curve, f1_score, matthews_corrcoef, jaccard_score, classification_report
# added
from sklearn.model_selection import train_test_split, cross_validate, cross_val_score, LeaveOneOut, KFold, RepeatedKFold, cross_val_predict
from sklearn.model_selection import train_test_split, cross_validate, cross_val_score
from sklearn.model_selection import StratifiedKFold,RepeatedStratifiedKFold, RepeatedKFold
from sklearn.pipeline import Pipeline, make_pipeline
from sklearn.feature_selection import RFE, RFECV
import itertools
import seaborn as sns
import matplotlib.pyplot as plt
from statistics import mean, stdev, median, mode
from imblearn.over_sampling import RandomOverSampler
from imblearn.under_sampling import RandomUnderSampler
from imblearn.over_sampling import SMOTE
from sklearn.datasets import make_classification
from imblearn.combine import SMOTEENN
from imblearn.combine import SMOTETomek
from imblearn.over_sampling import SMOTENC
from imblearn.under_sampling import EditedNearestNeighbours
from imblearn.under_sampling import RepeatedEditedNearestNeighbours
from sklearn.model_selection import GridSearchCV
from sklearn.base import BaseEstimator
from sklearn.impute import KNNImputer as KNN
import json
import argparse
import re
#%% GLOBALS
rs = {'random_state': 42}
njobs = {'n_jobs': 10}
scoring_fn = ({ 'mcc' : make_scorer(matthews_corrcoef)
, 'fscore' : make_scorer(f1_score)
, 'precision' : make_scorer(precision_score)
, 'recall' : make_scorer(recall_score)
, 'accuracy' : make_scorer(accuracy_score)
, 'roc_auc' : make_scorer(roc_auc_score)
, 'jcc' : make_scorer(jaccard_score)
})
skf_cv = StratifiedKFold(n_splits = 10
#, shuffle = False, random_state= None)
, shuffle = True,**rs)
rskf_cv = RepeatedStratifiedKFold(n_splits = 10
, n_repeats = 3
, **rs)
mcc_score_fn = {'mcc': make_scorer(matthews_corrcoef)}
jacc_score_fn = {'jcc': make_scorer(jaccard_score)}
###############################################################################
score_type_ordermapD = { 'mcc' : 1
, 'fscore' : 2
, 'jcc' : 3
, 'precision' : 4
, 'recall' : 5
, 'accuracy' : 6
, 'roc_auc' : 7
, 'TN' : 8
, 'FP' : 9
, 'FN' : 10
, 'TP' : 11
, 'trainingY_neg': 12
, 'trainingY_pos': 13
, 'blindY_neg' : 14
, 'blindY_pos' : 15
, 'fit_time' : 16
, 'score_time' : 17
}
scoreCV_mapD = {'test_mcc' : 'MCC'
, 'test_fscore' : 'F1'
, 'test_precision' : 'Precision'
, 'test_recall' : 'Recall'
, 'test_accuracy' : 'Accuracy'
, 'test_roc_auc' : 'ROC_AUC'
, 'test_jcc' : 'JCC'
}
scoreBT_mapD = {'bts_mcc' : 'MCC'
, 'bts_fscore' : 'F1'
, 'bts_precision' : 'Precision'
, 'bts_recall' : 'Recall'
, 'bts_accuracy' : 'Accuracy'
, 'bts_roc_auc' : 'ROC_AUC'
, 'bts_jcc' : 'JCC'
}
#%%############################################################################
############################
# MultModelsCl()
# Run Multiple Classifiers
############################
# Multiple Classification - Model Pipeline
def MultModelsCl(input_df, target, skf_cv
, blind_test_df
, blind_test_target
, tts_split_type
, resampling_type = 'none' # default
, add_cm = True # adds confusion matrix based on cross_val_predict
, add_yn = True # adds target var class numbers
, var_type = ['numerical', 'categorical','mixed']
, return_formatted_output = True):
'''
@ param input_df: input features
@ type: df with input features WITHOUT the target variable
@param target: target (or output) feature
@type: df or np.array or Series
@param skv_cv: stratifiedK fold int or object to allow shuffle and random state to pass
@type: int or StratifiedKfold()
@var_type: numerical, categorical and mixed to determine what col_transform to apply (MinMaxScalar and/or one-ho t encoder)
@type: list
returns
Dict containing multiple classification scores for each model and mean of each Stratified Kfold including training
'''
#======================================================
# Determine categorical and numerical features
#======================================================
numerical_ix = input_df.select_dtypes(include=['int64', 'float64']).columns
numerical_ix
categorical_ix = input_df.select_dtypes(include=['object', 'bool']).columns
categorical_ix
#======================================================
# Determine preprocessing steps ~ var_type
#======================================================
if var_type == 'numerical':
t = [('num', MinMaxScaler(), numerical_ix)]
if var_type == 'categorical':
t = [('cat', OneHotEncoder(), categorical_ix)]
if var_type == 'mixed':
t = [('num', MinMaxScaler(), numerical_ix)
, ('cat', OneHotEncoder(), categorical_ix) ]
col_transform = ColumnTransformer(transformers = t
, remainder='passthrough')
#======================================================
# Specify multiple Classification Models
#======================================================
models = [('AdaBoost Classifier' , AdaBoostClassifier(**rs) )
, ('Bagging Classifier' , BaggingClassifier(**rs, **njobs, bootstrap = True, oob_score = True) )
, ('Decision Tree' , DecisionTreeClassifier(**rs) )
, ('Extra Tree' , ExtraTreeClassifier(**rs) )
, ('Extra Trees' , ExtraTreesClassifier(**rs) )
, ('Gradient Boosting' , GradientBoostingClassifier(**rs) )
, ('Gaussian NB' , GaussianNB() )
, ('Gaussian Process' , GaussianProcessClassifier(**rs) )
, ('K-Nearest Neighbors' , KNeighborsClassifier() )
, ('LDA' , LinearDiscriminantAnalysis() )
, ('Logistic Regression' , LogisticRegression(**rs) )
, ('Logistic RegressionCV' , LogisticRegressionCV(cv = 3, **rs))
, ('MLP' , MLPClassifier(max_iter = 500, **rs) )
, ('Multinomial' , MultinomialNB() )
, ('Naive Bayes' , BernoulliNB() )
, ('Passive Aggresive' , PassiveAggressiveClassifier(**rs, **njobs) )
, ('QDA' , QuadraticDiscriminantAnalysis() )
, ('Random Forest' , RandomForestClassifier(**rs, n_estimators = 1000 ) )
, ('Random Forest2' , RandomForestClassifier(min_samples_leaf = 5
, n_estimators = 1000
, bootstrap = True
, oob_score = True
, **njobs
, **rs
, max_features = 'auto') )
, ('Ridge Classifier' , RidgeClassifier(**rs) )
, ('Ridge ClassifierCV' , RidgeClassifierCV(cv = 3) )
, ('SVC' , SVC(**rs) )
, ('Stochastic GDescent' , SGDClassifier(**rs, **njobs) )
, ('XGBoost' , XGBClassifier(**rs, verbosity = 0, use_label_encoder =False) )
]
mm_skf_scoresD = {}
print('\n==============================================================\n'
, '\nRunning several classification models (n):', len(models)
,'\nList of models:')
for m in models:
print(m)
print('\n================================================================\n')
index = 1
for model_name, model_fn in models:
print('\nRunning classifier:', index
, '\nModel_name:' , model_name
, '\nModel func:' , model_fn)
index = index+1
model_pipeline = Pipeline([
('prep' , col_transform)
, ('model' , model_fn)])
print('\nRunning model pipeline:', model_pipeline)
skf_cv_modD = cross_validate(model_pipeline
, input_df
, target
, cv = skf_cv
, scoring = scoring_fn
, return_train_score = True)
#######################################################################
#======================================================
# Option: Add confusion matrix from cross_val_predict
# Understand and USE with caution
# cross_val_score, cross_val_predict, "Passing these predictions into an evaluation metric may not be a valid way to measure generalization performance. Results can differ from cross_validate and cross_val_score unless all tests sets have equal size and the metric decomposes over samples."
# https://stackoverflow.com/questions/65645125/producing-a-confusion-matrix-with-cross-validate
#======================================================
if add_cm:
#-----------------------------------------------------------
# Initialise dict of Confusion Matrix (cm)
#-----------------------------------------------------------
cmD = {}
# Calculate cm
y_pred = cross_val_predict(model_pipeline, input_df, target, cv = skf_cv, **njobs)
#_tn, _fp, _fn, _tp = confusion_matrix(y_pred, y).ravel() # internally
tn, fp, fn, tp = confusion_matrix(y_pred, target).ravel()
# Build dict
cmD = {'TN' : tn
, 'FP': fp
, 'FN': fn
, 'TP': tp}
#---------------------------------
# Update cv dict with cmD and tbtD
#----------------------------------
skf_cv_modD.update(cmD)
else:
skf_cv_modD = skf_cv_modD
#######################################################################
#=============================================
# Option: Add targety numbers for data
#=============================================
if add_yn:
#-----------------------------------------------------------
# Initialise dict of target numbers: training and blind (tbt)
#-----------------------------------------------------------
tbtD = {}
# training y
tyn = Counter(target)
tyn_neg = tyn[0]
tyn_pos = tyn[1]
# blind test y
btyn = Counter(blind_test_target)
btyn_neg = btyn[0]
btyn_pos = btyn[1]
# Build dict
tbtD = {'n_trainingY_neg' : tyn_neg
, 'n_trainingY_pos' : tyn_pos
, 'n_blindY_neg' : btyn_neg
, 'n_blindY_pos' : btyn_pos}
#---------------------------------
# Update cv dict with cmD and tbtD
#----------------------------------
skf_cv_modD.update(tbtD)
else:
skf_cv_modD = skf_cv_modD
#######################################################################
#==============================
# Extract mean values for CV
#==============================
mm_skf_scoresD[model_name] = {}
for key, value in skf_cv_modD.items():
print('\nkey:', key, '\nvalue:', value)
print('\nmean value:', np.mean(value))
mm_skf_scoresD[model_name][key] = round(np.mean(value),2)
#return(mm_skf_scoresD)
#%%
#=========================
# Blind test: BTS results
#=========================
# Build the final results with all scores for the model
#bts_predict = gscv_fs.predict(blind_test_df)
model_pipeline.fit(input_df, target)
bts_predict = model_pipeline.predict(blind_test_df)
bts_mcc_score = round(matthews_corrcoef(blind_test_target, bts_predict),2)
print('\nMCC on Blind test:' , bts_mcc_score)
print('\nAccuracy on Blind test:', round(accuracy_score(blind_test_target, bts_predict),2))
# Diff b/w train and bts test scores
# train_test_diff_MCC = cvtrain_mcc - bts_mcc_score
# print('\nDiff b/w train and blind test score (MCC):', train_test_diff)
mm_skf_scoresD[model_name]['bts_mcc'] = bts_mcc_score
mm_skf_scoresD[model_name]['bts_fscore'] = round(f1_score(blind_test_target, bts_predict),2)
mm_skf_scoresD[model_name]['bts_precision'] = round(precision_score(blind_test_target, bts_predict),2)
mm_skf_scoresD[model_name]['bts_recall'] = round(recall_score(blind_test_target, bts_predict),2)
mm_skf_scoresD[model_name]['bts_accuracy'] = round(accuracy_score(blind_test_target, bts_predict),2)
mm_skf_scoresD[model_name]['bts_roc_auc'] = round(roc_auc_score(blind_test_target, bts_predict),2)
mm_skf_scoresD[model_name]['bts_jcc'] = round(jaccard_score(blind_test_target, bts_predict),2)
#mm_skf_scoresD[model_name]['diff_mcc'] = train_test_diff_MCC
#return(mm_skf_scoresD)
#%%
# ADD more info: meta data related to input and blind and resampling
# target numbers: training
yc1 = Counter(target)
yc1_ratio = yc1[0]/yc1[1]
# target numbers: test
yc2 = Counter(blind_test_target)
yc2_ratio = yc2[0]/yc2[1]
mm_skf_scoresD[model_name]['resampling'] = resampling_type
mm_skf_scoresD[model_name]['n_training_size'] = len(input_df)
mm_skf_scoresD[model_name]['n_trainingY_ratio'] = round(yc1_ratio, 2)
mm_skf_scoresD[model_name]['n_test_size'] = len(blind_test_df)
mm_skf_scoresD[model_name]['n_testY_ratio'] = round(yc2_ratio,2)
mm_skf_scoresD[model_name]['n_features'] = len(input_df.columns)
mm_skf_scoresD[model_name]['tts_split'] = tts_split_type
#return(mm_skf_scoresD)
#============================
# Process the dict to have WF
#============================
if return_formatted_output:
CV_BT_metaDF = ProcessMultModelsCl(mm_skf_scoresD)
return(CV_BT_metaDF)
else:
return(mm_skf_scoresD)
#%% Process output function ###################################################
############################
# ProcessMultModelsCl()
############################
#Processes the dict from above if use_formatted_output = True
def ProcessMultModelsCl(inputD = {}):
scoresDF = pd.DataFrame(inputD)
#------------------------
# Extracting split_name
#-----------------------
tts_split_nameL = []
for k,v in inputD.items():
tts_split_nameL = tts_split_nameL + [v['tts_split']]
if len(set(tts_split_nameL)) == 1:
tts_split_name = str(list(set(tts_split_nameL))[0])
print('\nExtracting tts_split_name:', tts_split_name)
#------------------------
# WF: only CV and BTS
#-----------------------
scoresDFT = scoresDF.T
scoresDF_CV = scoresDFT.filter(regex='^test_.*$', axis = 1); scoresDF_CV.columns
# map colnames for consistency to allow concatenting
scoresDF_CV.columns = scoresDF_CV.columns.map(scoreCV_mapD); scoresDF_CV.columns
scoresDF_CV['source_data'] = 'CV'
scoresDF_BT = scoresDFT.filter(regex='^bts_.*$', axis = 1); scoresDF_BT.columns
# map colnames for consistency to allow concatenting
scoresDF_BT.columns = scoresDF_BT.columns.map(scoreBT_mapD); scoresDF_BT.columns
scoresDF_BT['source_data'] = 'BT'
# dfs_combine_wf = [baseline_BT, smnc_BT, ros_BT, rus_BT, rouC_BT,
# baseline_CV, smnc_CV, ros_CV, rus_CV, rouC_CV]
#baseline_all = baseline_all_scores.filter(regex = 'bts_.*|test_.*|.*_time|TN|FP|FN|TP|.*_neg|.*_pos', axis = 0)
#metaDF = scoresDFT.filter(regex='training_size|blind_test_size|_time|TN|FP|FN|TP|.*_neg|.*_pos|resampling', axis = 1); scoresDF_BT.columns
#metaDF = scoresDFT.filter(regex='n_.*$|_time|TN|FP|FN|TP|.*_neg|.*_pos|resampling|tts.*', axis = 1); metaDF.columns
metaDF = scoresDFT.filter(regex='^(?!test_.*$|bts_.*$|train_.*$).*'); metaDF.columns
print('\nTotal cols in each df:'
, '\nCV df:', len(scoresDF_CV.columns)
, '\nBT_df:', len(scoresDF_BT.columns)
, '\nmetaDF:', len(metaDF.columns))
if len(scoresDF_CV.columns) == len(scoresDF_BT.columns):
print('\nFirst proceeding to rowbind CV and BT dfs:')
expected_ncols_out = len(scoresDF_BT.columns) + len(metaDF.columns)
print('\nFinal output should have:', expected_ncols_out, 'columns' )
#-----------------
# Combine WF
#-----------------
dfs_combine_wf = [scoresDF_CV, scoresDF_BT]
print('\nCombinig', len(dfs_combine_wf), 'using pd.concat by row ~ rowbind'
, '\nChecking Dims of df to combine:'
, '\nDim of CV:', scoresDF_CV.shape
, '\nDim of BT:', scoresDF_BT.shape)
#print(scoresDF_CV)
#print(scoresDF_BT)
dfs_nrows_wf = []
for df in dfs_combine_wf:
dfs_nrows_wf = dfs_nrows_wf + [len(df)]
dfs_nrows_wf = max(dfs_nrows_wf)
dfs_ncols_wf = []
for df in dfs_combine_wf:
dfs_ncols_wf = dfs_ncols_wf + [len(df.columns)]
dfs_ncols_wf = max(dfs_ncols_wf)
print(dfs_ncols_wf)
expected_nrows_wf = len(dfs_combine_wf) * dfs_nrows_wf
expected_ncols_wf = dfs_ncols_wf
common_cols_wf = list(set.intersection(*(set(df.columns) for df in dfs_combine_wf)))
print('\nNumber of Common columns:', dfs_ncols_wf
, '\nThese are:', common_cols_wf)
if len(common_cols_wf) == dfs_ncols_wf :
combined_baseline_wf = pd.concat([df[common_cols_wf] for df in dfs_combine_wf], ignore_index=False)
print('\nConcatenating dfs with different resampling methods [WF]:'
, '\nSplit type:', tts_split_name
, '\nNo. of dfs combining:', len(dfs_combine_wf))
#print('\n================================================^^^^^^^^^^^^')
if len(combined_baseline_wf) == expected_nrows_wf and len(combined_baseline_wf.columns) == expected_ncols_wf:
#print('\n================================================^^^^^^^^^^^^')
print('\nPASS:', len(dfs_combine_wf), 'dfs successfully combined'
, '\nnrows in combined_df_wf:', len(combined_baseline_wf)
, '\nncols in combined_df_wf:', len(combined_baseline_wf.columns))
else:
print('\nFAIL: concatenating failed'
, '\nExpected nrows:', expected_nrows_wf
, '\nGot:', len(combined_baseline_wf)
, '\nExpected ncols:', expected_ncols_wf
, '\nGot:', len(combined_baseline_wf.columns))
sys.exit('\nFIRST IF FAILS')
else:
print('\nConcatenting dfs not possible [WF],check numbers ')
#-------------------------------------
# Combine WF+Metadata: Final output
#-------------------------------------
# checking indices for the dfs to combine:
c1L = list(set(combined_baseline_wf.index))
c2L = list(metaDF.index)
#if set(c1L) == set(c2L):
if set(c1L) == set(c2L) and all(x in c2L for x in c1L) and all(x in c1L for x in c2L):
print('\nPASS: proceeding to merge metadata with CV and BT dfs')
combDF = pd.merge(combined_baseline_wf, metaDF, left_index = True, right_index = True)
else:
sys.exit('\nFAIL: Could not merge metadata with CV and BT dfs')
if len(combDF.columns) == expected_ncols_out:
print('\nPASS: Combined df has expected ncols')
else:
sys.exit('\nFAIL: Length mismatch for combined_df')
print('\nAdding column: Model_name')
combDF['Model_name'] = combDF.index
print('\n========================================================='
, '\nSUCCESS: Ran multiple classifiers'
, '\n=======================================================')
#resampling_methods_wf = combined_baseline_wf[['resampling']]
#resampling_methods_wf = resampling_methods_wf.drop_duplicates()
#, '\n', resampling_methods_wf)
return combDF
###############################################################################
#%% Feature selection function ################################################
############################
# fsgs_rfecv()
############################
# Run FS using some classifier models
#
def fsgs_rfecv(input_df
, target
, param_gridLd = [{'fs__min_features_to_select' : [1]}]
, blind_test_df = pd.DataFrame()
, blind_test_target = pd.Series(dtype = 'int64')
, estimator = LogisticRegression(**rs) # placeholder
, use_fs = False # uses estimator as the RFECV parameter for fs. Set to TRUE if you want to supply custom_fs as shown below
, custom_fs = RFECV(DecisionTreeClassifier(**rs) , cv = skf_cv, scoring = 'matthews_corrcoef')
, cv_method = skf_cv
, var_type = ['numerical', 'categorical' , 'mixed']
, verbose = 3
):
'''
returns
Dict containing results from FS and hyperparam tuning for a given estiamtor
>>> ADD MORE <<<
optimised/selected based on mcc
'''
###########################################################################
#================================================
# Determine categorical and numerical features
#================================================
numerical_ix = input_df.select_dtypes(include=['int64', 'float64']).columns
numerical_ix
categorical_ix = input_df.select_dtypes(include=['object', 'bool']).columns
categorical_ix
#================================================
# Determine preprocessing steps ~ var_type
#================================================
if var_type == 'numerical':
t = [('num', MinMaxScaler(), numerical_ix)]
if var_type == 'categorical':
t = [('cat', OneHotEncoder(), categorical_ix)]
if var_type == 'mixed':
t = [('cat', OneHotEncoder(), categorical_ix)
, ('num', MinMaxScaler(), numerical_ix)]
col_transform = ColumnTransformer(transformers = t
, remainder='passthrough')
###########################################################################
#==================================================
# Create var_type ~ column names
# using one hot encoder with RFECV means
# the names internally are lost. Hence
# fit col_transformeer to my input_df and get
# all the column names out and stored in a var
# to allow the 'selected features' to be subsetted
# from the numpy boolean array
#=================================================
col_transform.fit(input_df)
col_transform.get_feature_names_out()
var_type_colnames = col_transform.get_feature_names_out()
var_type_colnames = pd.Index(var_type_colnames)
if var_type == 'mixed':
print('\nVariable type is:', var_type
, '\nNo. of columns in input_df:', len(input_df.columns)
, '\nNo. of columns post one hot encoder:', len(var_type_colnames))
else:
print('\nNo. of columns in input_df:', len(input_df.columns))
#==================================
# Build FS with supplied estimator
#==================================
if use_fs:
fs = custom_fs
else:
fs = RFECV(estimator, cv = skf_cv, scoring = 'matthews_corrcoef')
#==================================
# Build basic param grid
#==================================
# param_gridD = [
# {'fs__min_features_to_select' : [1]
# }]
############################################################################
# Create Pipeline object
pipe = Pipeline([
('pre', col_transform),
('fs', fs),
('clf', estimator)])
############################################################################
# Define GridSearchCV
gscv_fs = GridSearchCV(pipe
#, param_gridLd = param_gridD
, param_gridLd
, cv = cv_method
, scoring = scoring_fn
, refit = 'mcc'
, verbose = 3
, return_train_score = True
, **njobs)
gscv_fs.fit(input_df, target)
###########################################################################
# Get best param and scores out
gscv_fs.best_params_
gscv_fs.best_score_
# Training best score corresponds to the max of the mean_test<score>
train_bscore = round(gscv_fs.best_score_, 2); train_bscore
print('\nTraining best score (MCC):', train_bscore)
gscv_fs.cv_results_['mean_test_mcc']
round(gscv_fs.cv_results_['mean_test_mcc'].max(),2)
round(np.nanmax(gscv_fs.cv_results_['mean_test_mcc']),2)
check_train_score = [round(gscv_fs.cv_results_['mean_test_mcc'].max(),2)
, round(np.nanmax(gscv_fs.cv_results_['mean_test_mcc']),2)]
check_train_score = np.nanmax(check_train_score)
# Training results
gscv_tr_resD = gscv_fs.cv_results_
mod_refit_param = gscv_fs.refit
# sanity check
if train_bscore == check_train_score:
print('\nVerified training score (MCC):', train_bscore )
else:
sys.exit('\nTraining score could not be internatlly verified. Please check training results dict')
#-------------------------
# Dict of CV results
#-------------------------
cv_allD = gscv_fs.cv_results_
cvdf0 = pd.DataFrame(cv_allD)
cvdf = cvdf0.filter(regex='mean_test', axis = 1)
cvdfT = cvdf.T
cvdfT.columns = ['cv_score']
cvdfTr = cvdfT.loc[:,'cv_score'].round(decimals = 2) # round values
cvD = cvdfTr.to_dict()
print('\n CV results dict generated for:', len(scoring_fn), 'scores'
, '\nThese are:', scoring_fn.keys())
#-------------------------
# Blind test: REAL check!
#-------------------------
#tp = gscv_fs.predict(X_bts)
tp = gscv_fs.predict(blind_test_df)
print('\nMCC on Blind test:' , round(matthews_corrcoef(blind_test_target, tp),2))
print('\nAccuracy on Blind test:', round(accuracy_score(blind_test_target, tp),2))
#=================
# info extraction
#=================
# gives input vals??
gscv_fs._check_n_features
# gives gscv params used
gscv_fs._get_param_names()
# gives ??
gscv_fs.best_estimator_
gscv_fs.best_params_ # gives best estimator params as a dict
gscv_fs.best_estimator_._final_estimator # similar to above, doesn't contain max_iter
gscv_fs.best_estimator_.named_steps['fs'].get_support()
gscv_fs.best_estimator_.named_steps['fs'].ranking_ # array of ranks for the features
gscv_fs.best_estimator_.named_steps['fs'].grid_scores_.mean()
gscv_fs.best_estimator_.named_steps['fs'].grid_scores_.max()
#gscv_fs.best_estimator_.named_steps['fs'].grid_scores_
estimator_mask = gscv_fs.best_estimator_.named_steps['fs'].get_support()
############################################################################
#============
# FS results
#============
# Now get the features out
#--------------
# All features
#--------------
all_features = gscv_fs.feature_names_in_
n_all_features = gscv_fs.n_features_in_
#all_features = gsfit.feature_names_in_
#--------------
# Selected features by the classifier
# Important to have var_type_colnames here
#----------------
#sel_features = X.columns[gscv_fs.best_estimator_.named_steps['fs'].get_support()] 3 only for numerical df
sel_features = var_type_colnames[gscv_fs.best_estimator_.named_steps['fs'].get_support()]
n_sf = gscv_fs.best_estimator_.named_steps['fs'].n_features_
#--------------
# Get model name
#--------------
model_name = gscv_fs.best_estimator_.named_steps['clf']
b_model_params = gscv_fs.best_params_
print('\n========================================'
, '\nRunning model:'
, '\nModel name:', model_name
, '\n==============================================='
, '\nRunning feature selection with RFECV for model'
, '\nTotal no. of features in model:', len(all_features)
, '\nThese are:\n', all_features, '\n\n'
, '\nNo of features for best model: ', n_sf
, '\nThese are:', sel_features, '\n\n'
, '\nBest Model hyperparams:', b_model_params
)
###########################################################################
############################## OUTPUT #####################################
###########################################################################
#=========================
# Blind test: BTS results
#=========================
# Build the final results with all scores for a feature selected model
#bts_predict = gscv_fs.predict(X_bts)
bts_predict = gscv_fs.predict(blind_test_df)
print('\nMCC on Blind test:' , round(matthews_corrcoef(blind_test_target, bts_predict),2))
print('\nAccuracy on Blind test:', round(accuracy_score(blind_test_target, bts_predict),2))
bts_mcc_score = round(matthews_corrcoef(blind_test_target, bts_predict),2)
# Diff b/w train and bts test scores
train_test_diff = train_bscore - bts_mcc_score
print('\nDiff b/w train and blind test score (MCC):', train_test_diff)
lr_btsD ={}
#lr_btsD['bts_mcc'] = bts_mcc_score
lr_btsD['bts_fscore'] = round(f1_score(blind_test_target, bts_predict),2)
lr_btsD['bts_precision'] = round(precision_score(blind_test_target, bts_predict),2)
lr_btsD['bts_recall'] = round(recall_score(blind_test_target, bts_predict),2)
lr_btsD['bts_accuracy'] = round(accuracy_score(blind_test_target, bts_predict),2)
lr_btsD['bts_roc_auc'] = round(roc_auc_score(blind_test_target, bts_predict),2)
lr_btsD['bts_jcc'] = round(jaccard_score(blind_test_target, bts_predict),2)
lr_btsD
#===========================
# Add FS related model info
#===========================
model_namef = str(model_name)
# FIXME: doesn't tell you which it has chosen
fs_methodf = str(gscv_fs.best_estimator_.named_steps['fs'])
all_featuresL = list(all_features)
fs_res_arrayf = str(list( gscv_fs.best_estimator_.named_steps['fs'].get_support()))
fs_res_array_rankf = str(list( gscv_fs.best_estimator_.named_steps['fs'].ranking_))
sel_featuresf = list(sel_features)
n_sf = int(n_sf)
output_modelD = {'model_name': model_namef
, 'model_refit_param': mod_refit_param
, 'Best_model_params': b_model_params
, 'n_all_features': n_all_features
, 'fs_method': fs_methodf
, 'fs_res_array': fs_res_arrayf
, 'fs_res_array_rank': fs_res_array_rankf
, 'all_feature_names': all_featuresL
, 'n_sel_features': n_sf
, 'sel_features_names': sel_featuresf}
#output_modelD
#========================================
# Update output_modelD with bts_results
#========================================
output_modelD.update(lr_btsD)
output_modelD
output_modelD['train_score (MCC)'] = train_bscore
output_modelD['bts_mcc'] = bts_mcc_score
output_modelD['train_bts_diff'] = round(train_test_diff,2)
print(output_modelD)
nlen = len(output_modelD)
#========================================
# Update output_modelD with cv_results
#========================================
output_modelD.update(cvD)
if (len(output_modelD) == nlen + len(cvD)):
print('\nFS run complete for model:', estimator
, '\nFS using:', fs
, '\nOutput dict size:', len(output_modelD))
return(output_modelD)
else:
sys.exit('\nFAIL:numbers mismatch output dict length not as expected. Please check')

595033
scripts/ml/log_FS_pnca_7030.txt
View file

File diff suppressed because one or more lines are too long

730
scripts/ml/ml_data.py
View file

@ -1,730 +0,0 @@
#!/usr/bin/env python3
# -*- coding: utf-8 -*-
"""
Created on Sun Mar 6 13:41:54 2022
@author: tanu
"""
#def setvars(gene,drug):
#https://stackoverflow.com/questions/51695322/compare-multiple-algorithms-with-sklearn-pipeline
import os, sys
import pandas as pd
import numpy as np
print(np.__version__)
print(pd.__version__)
import pprint as pp
from copy import deepcopy
from collections import Counter
from sklearn.impute import KNNImputer as KNN
from imblearn.over_sampling import RandomOverSampler
from imblearn.under_sampling import RandomUnderSampler
from imblearn.over_sampling import SMOTE
from sklearn.datasets import make_classification
from imblearn.combine import SMOTEENN
from imblearn.combine import SMOTETomek
from imblearn.over_sampling import SMOTENC
from imblearn.under_sampling import EditedNearestNeighbours
from imblearn.under_sampling import RepeatedEditedNearestNeighbours
from sklearn.metrics import make_scorer, confusion_matrix, accuracy_score, balanced_accuracy_score, precision_score, average_precision_score, recall_score
from sklearn.metrics import roc_auc_score, roc_curve, f1_score, matthews_corrcoef, jaccard_score, classification_report
from sklearn.model_selection import train_test_split, cross_validate, cross_val_score
from sklearn.model_selection import StratifiedKFold,RepeatedStratifiedKFold, RepeatedKFold
from sklearn.pipeline import Pipeline, make_pipeline
#%% GLOBALS
rs = {'random_state': 42}
njobs = {'n_jobs': 10}
scoring_fn = ({ 'mcc' : make_scorer(matthews_corrcoef)
, 'accuracy' : make_scorer(accuracy_score)
, 'fscore' : make_scorer(f1_score)
, 'precision' : make_scorer(precision_score)
, 'recall' : make_scorer(recall_score)
, 'roc_auc' : make_scorer(roc_auc_score)
, 'jcc' : make_scorer(jaccard_score)
})
skf_cv = StratifiedKFold(n_splits = 10
#, shuffle = False, random_state= None)
, shuffle = True,**rs)
rskf_cv = RepeatedStratifiedKFold(n_splits = 10
, n_repeats = 3
, **rs)
mcc_score_fn = {'mcc': make_scorer(matthews_corrcoef)}
jacc_score_fn = {'jcc': make_scorer(jaccard_score)}
#%% FOR LATER: Combine ED logo data
#%% DONE: active aa site annotations **DONE on 15/05/2022 as part of generating merged_dfs
###########################################################################
rs = {'random_state': 42}
njobs = {'n_jobs': 10}
homedir = os.path.expanduser("~")
geneL_basic = ['pnca']
geneL_na = ['gid']
geneL_na_ppi2 = ['rpob']
geneL_ppi2 = ['alr', 'embb', 'katg']
#num_type = ['int64', 'float64']
num_type = ['int16', 'int32', 'int64', 'float16', 'float32', 'float64']
cat_type = ['object', 'bool']
#==============
# directories
#==============
datadir = homedir + '/git/Data/'
indir = datadir + drug + '/input/'
outdir = datadir + drug + '/output/'
#=======
# input
#=======
#---------
# File 1
#---------
infile_ml1 = outdir + gene.lower() + '_merged_df3.csv'
#infile_ml2 = outdir + gene.lower() + '_merged_df2.csv'
my_features_df = pd.read_csv(infile_ml1, index_col = 0)
my_features_df = my_features_df .reset_index(drop = True)
my_features_df.index
my_features_df.dtypes
mycols = my_features_df.columns
#---------
# File 2
#---------
infile_aaindex = outdir + 'aa_index/' + gene.lower() + '_aa.csv'
aaindex_df = pd.read_csv(infile_aaindex, index_col = 0)
aaindex_df.dtypes
#-----------
# check for non-numerical columns
#-----------
if any(aaindex_df.dtypes==object):
print('\naaindex_df contains non-numerical data')
aaindex_df_object = aaindex_df.select_dtypes(include = cat_type)
print('\nTotal no. of non-numerial columns:', len(aaindex_df_object.columns))
expected_aa_ncols = len(aaindex_df.columns) - len(aaindex_df_object.columns)
#-----------
# Extract numerical data only
#-----------
print('\nSelecting numerical data only')
aaindex_df = aaindex_df.select_dtypes(include = num_type)
#---------------------------
# aaindex: sanity check 1
#---------------------------
if len(aaindex_df.columns) == expected_aa_ncols:
print('\nPASS: successfully selected numerical columns only for aaindex_df')
else:
print('\nFAIL: Numbers mismatch'
, '\nExpected ncols:', expected_aa_ncols
, '\nGot:', len(aaindex_df.columns))
#---------------
# check for NA
#---------------
print('\nNow checking for NA in the remaining aaindex_cols')
c1 = aaindex_df.isna().sum()
c2 = c1.sort_values(ascending=False)
print('\nCounting aaindex_df cols with NA'
, '\nncols with NA:', sum(c2>0), 'columns'
, '\nDropping these...'
, '\nOriginal ncols:', len(aaindex_df.columns)
)
aa_df = aaindex_df.dropna(axis=1)
print('\nRevised df ncols:', len(aa_df.columns))
c3 = aa_df.isna().sum()
c4 = c3.sort_values(ascending=False)
print('\nChecking NA in revised df...')
if sum(c4>0):
sys.exit('\nFAIL: aaindex_df still contains cols with NA, please check and drop these before proceeding...')
else:
print('\nPASS: cols with NA successfully dropped from aaindex_df'
, '\nProceeding with combining aa_df with other features_df')
#---------------------------
# aaindex: sanity check 2
#---------------------------
expected_aa_ncols2 = len(aaindex_df.columns) - sum(c2>0)
if len(aa_df.columns) == expected_aa_ncols2:
print('\nPASS: ncols match'
, '\nExpected ncols:', expected_aa_ncols2
, '\nGot:', len(aa_df.columns))
else:
print('\nFAIL: Numbers mismatch'
, '\nExpected ncols:', expected_aa_ncols2
, '\nGot:', len(aa_df.columns))
# Important: need this to identify aaindex cols
aa_df_cols = aa_df.columns
print('\nTotal no. of columns in clean aa_df:', len(aa_df_cols))
###############################################################################
#%% Combining my_features_df and aaindex_df
#===========================
# Merge my_df + aaindex_df
#===========================
if aa_df.columns[aa_df.columns.isin(my_features_df.columns)] == my_features_df.columns[my_features_df.columns.isin(aa_df.columns)]:
print('\nMerging on column: mutationinformation')
if len(my_features_df) == len(aa_df):
expected_nrows = len(my_features_df)
print('\nProceeding to merge, expected nrows in merged_df:', expected_nrows)
else:
sys.exit('\nNrows mismatch, cannot merge. Please check'
, '\nnrows my_df:', len(my_features_df)
, '\nnrows aa_df:', len(aa_df))
#-----------------
# Reset index: mutationinformation
# Very important for merging
#-----------------
aa_df = aa_df.reset_index()
expected_ncols = len(my_features_df.columns) + len(aa_df.columns) - 1 # for the no. of merging col
#-----------------
# Merge: my_features_df + aa_df
#-----------------
merged_df = pd.merge(my_features_df
, aa_df
, on = 'mutationinformation')
#---------------------------
# aaindex: sanity check 3
#---------------------------
if len(merged_df.columns) == expected_ncols:
print('\nPASS: my_features_df and aa_df successfully combined'
, '\nnrows:', len(merged_df)
, '\nncols:', len(merged_df.columns))
else:
sys.exit('\nFAIL: could not combine my_features_df and aa_df'
, '\nCheck dims and merging cols!')
#--------
# Reassign so downstream code doesn't need to change
#--------
my_df = merged_df.copy()
#%% Data: my_df
# Check if non structural pos have crept in
# IDEALLY remove from source! But for rpoB do it here
# Drop NA where numerical cols have them
if gene.lower() in geneL_na_ppi2:
#D1148 get rid of
na_index = my_df['mutationinformation'].index[my_df['mcsm_na_affinity'].apply(np.isnan)]
my_df = my_df.drop(index=na_index)
# FIXED: complete data for all muts inc L114M, F115L, V123L, V125I, V131M
# if gene.lower() in ['embb']:
# na_index = my_df['mutationinformation'].index[my_df['ligand_distance'].apply(np.isnan)]
# my_df = my_df.drop(index=na_index)
# # Sanity check for non-structural positions
# print('\nChecking for non-structural postions')
# na_index = my_df['mutationinformation'].index[my_df['ligand_distance'].apply(np.isnan)]
# if len(na_index) > 0:
# print('\nNon-structural positions detected for gene:', gene.lower()
# , '\nTotal number of these detected:', len(na_index)
# , '\These are at index:', na_index
# , '\nOriginal nrows:', len(my_df)
# , '\nDropping these...')
# my_df = my_df.drop(index=na_index)
# print('\nRevised nrows:', len(my_df))
# else:
# print('\nNo non-structural positions detected for gene:', gene.lower()
# , '\nnrows:', len(my_df))
###########################################################################
#%% Add lineage calculation columns
#FIXME: Check if this can be imported from config?
total_mtblineage_uc = 8
lineage_colnames = ['lineage_list_all', 'lineage_count_all', 'lineage_count_unique', 'lineage_list_unique', 'lineage_multimode']
#bar = my_df[lineage_colnames]
my_df['lineage_proportion'] = my_df['lineage_count_unique']/my_df['lineage_count_all']
my_df['dist_lineage_proportion'] = my_df['lineage_count_unique']/total_mtblineage_uc
###########################################################################
#%% Active site annotation column
# change from numberic to categorical
if my_df['active_site'].dtype in num_type:
my_df['active_site'] = my_df['active_site'].astype(object)
my_df['active_site'].dtype
#%% AA property change
#--------------------
# Water prop change
#--------------------
my_df['water_change'] = my_df['wt_prop_water'] + str('_to_') + my_df['mut_prop_water']
my_df['water_change'].value_counts()
water_prop_changeD = {
'hydrophobic_to_neutral' : 'change'
, 'hydrophobic_to_hydrophobic' : 'no_change'
, 'neutral_to_neutral' : 'no_change'
, 'neutral_to_hydrophobic' : 'change'
, 'hydrophobic_to_hydrophilic' : 'change'
, 'neutral_to_hydrophilic' : 'change'
, 'hydrophilic_to_neutral' : 'change'
, 'hydrophilic_to_hydrophobic' : 'change'
, 'hydrophilic_to_hydrophilic' : 'no_change'
}
my_df['water_change'] = my_df['water_change'].map(water_prop_changeD)
my_df['water_change'].value_counts()
#--------------------
# Polarity change
#--------------------
my_df['polarity_change'] = my_df['wt_prop_polarity'] + str('_to_') + my_df['mut_prop_polarity']
my_df['polarity_change'].value_counts()
polarity_prop_changeD = {
'non-polar_to_non-polar' : 'no_change'
, 'non-polar_to_neutral' : 'change'
, 'neutral_to_non-polar' : 'change'
, 'neutral_to_neutral' : 'no_change'
, 'non-polar_to_basic' : 'change'
, 'acidic_to_neutral' : 'change'
, 'basic_to_neutral' : 'change'
, 'non-polar_to_acidic' : 'change'
, 'neutral_to_basic' : 'change'
, 'acidic_to_non-polar' : 'change'
, 'basic_to_non-polar' : 'change'
, 'neutral_to_acidic' : 'change'
, 'acidic_to_acidic' : 'no_change'
, 'basic_to_acidic' : 'change'
, 'basic_to_basic' : 'no_change'
, 'acidic_to_basic' : 'change'}
my_df['polarity_change'] = my_df['polarity_change'].map(polarity_prop_changeD)
my_df['polarity_change'].value_counts()
#--------------------
# Electrostatics change
#--------------------
my_df['electrostatics_change'] = my_df['wt_calcprop'] + str('_to_') + my_df['mut_calcprop']
my_df['electrostatics_change'].value_counts()
calc_prop_changeD = {
'non-polar_to_non-polar' : 'no_change'
, 'non-polar_to_polar' : 'change'
, 'polar_to_non-polar' : 'change'
, 'non-polar_to_pos' : 'change'
, 'neg_to_non-polar' : 'change'
, 'non-polar_to_neg' : 'change'
, 'pos_to_polar' : 'change'
, 'pos_to_non-polar' : 'change'
, 'polar_to_polar' : 'no_change'
, 'neg_to_neg' : 'no_change'
, 'polar_to_neg' : 'change'
, 'pos_to_neg' : 'change'
, 'pos_to_pos' : 'no_change'
, 'polar_to_pos' : 'change'
, 'neg_to_polar' : 'change'
, 'neg_to_pos' : 'change'
}
my_df['electrostatics_change'] = my_df['electrostatics_change'].map(calc_prop_changeD)
my_df['electrostatics_change'].value_counts()
#--------------------
# Summary change: Create a combined column summarising these three cols
#--------------------
detect_change = 'change'
check_prop_cols = ['water_change', 'polarity_change', 'electrostatics_change']
#my_df['aa_prop_change'] = (my_df.values == detect_change).any(1).astype(int)
my_df['aa_prop_change'] = (my_df[check_prop_cols].values == detect_change).any(1).astype(int)
my_df['aa_prop_change'].value_counts()
my_df['aa_prop_change'].dtype
my_df['aa_prop_change'] = my_df['aa_prop_change'].map({1:'change'
, 0: 'no_change'})
my_df['aa_prop_change'].value_counts()
my_df['aa_prop_change'].dtype
#%% IMPUTE values for OR [check script for exploration: UQ_or_imputer]
#--------------------
# Impute OR values
#--------------------
#or_cols = ['or_mychisq', 'log10_or_mychisq', 'or_fisher']
sel_cols = ['mutationinformation', 'or_mychisq', 'log10_or_mychisq']
or_cols = ['or_mychisq', 'log10_or_mychisq']
print("count of NULL values before imputation\n")
print(my_df[or_cols].isnull().sum())
my_dfI = pd.DataFrame(index = my_df['mutationinformation'] )
my_dfI = pd.DataFrame(KNN(n_neighbors=3, weights="uniform").fit_transform(my_df[or_cols])
, index = my_df['mutationinformation']
, columns = or_cols )
my_dfI.columns = ['or_rawI', 'logorI']
my_dfI.columns
my_dfI = my_dfI.reset_index(drop = False) # prevents old index from being added as a column
my_dfI.head()
print("count of NULL values AFTER imputation\n")
print(my_dfI.isnull().sum())
#-------------------------------------------
# OR df Merge: with original based on index
#-------------------------------------------
#my_df['index_bm'] = my_df.index
mydf_imputed = pd.merge(my_df
, my_dfI
, on = 'mutationinformation')
#mydf_imputed = mydf_imputed.set_index(['index_bm'])
my_df['log10_or_mychisq'].isna().sum()
mydf_imputed['log10_or_mychisq'].isna().sum()
mydf_imputed['logorI'].isna().sum() # should be 0
len(my_df.columns)
len(mydf_imputed.columns)
#-----------------------------------------
# REASSIGN my_df after imputing OR values
#-----------------------------------------
my_df = mydf_imputed.copy()
if my_df['logorI'].isna().sum() == 0:
print('\nPASS: OR values imputed, data ready for ML')
else:
sys.exit('\nFAIL: something went wrong, Data not ready for ML. Please check upstream!')
#!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
#---------------------------------------
# TODO: try other imputation like MICE
#---------------------------------------
#!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
#%%########################################################################
#==========================
# Data for ML
#==========================
my_df_ml = my_df.copy()
#%% Build X: input for ML
common_cols_stabiltyN = ['ligand_distance'
, 'ligand_affinity_change'
, 'duet_stability_change'
, 'ddg_foldx'
, 'deepddg'
, 'ddg_dynamut2'
, 'mmcsm_lig'
, 'contacts']
# Build stability columns ~ gene
if gene.lower() in geneL_basic:
X_stabilityN = common_cols_stabiltyN
cols_to_mask = ['ligand_affinity_change']
if gene.lower() in geneL_ppi2:
# X_stabilityN = common_cols_stabiltyN + ['mcsm_ppi2_affinity' , 'interface_dist']
geneL_ppi2_st_cols = ['mcsm_ppi2_affinity', 'interface_dist']
X_stabilityN = common_cols_stabiltyN + geneL_ppi2_st_cols
cols_to_mask = ['ligand_affinity_change', 'mcsm_ppi2_affinity']
if gene.lower() in geneL_na:
# X_stabilityN = common_cols_stabiltyN + ['mcsm_na_affinity']
geneL_na_st_cols = ['mcsm_na_affinity']
X_stabilityN = common_cols_stabiltyN + geneL_na_st_cols
cols_to_mask = ['ligand_affinity_change', 'mcsm_na_affinity']
if gene.lower() in geneL_na_ppi2:
# X_stabilityN = common_cols_stabiltyN + ['mcsm_na_affinity'] + ['mcsm_ppi2_affinity', 'interface_dist']
geneL_na_ppi2_st_cols = ['mcsm_na_affinity'] + ['mcsm_ppi2_affinity', 'interface_dist']
X_stabilityN = common_cols_stabiltyN + geneL_na_ppi2_st_cols
cols_to_mask = ['ligand_affinity_change', 'mcsm_na_affinity', 'mcsm_ppi2_affinity']
X_foldX_cols = [ 'electro_rr', 'electro_mm', 'electro_sm', 'electro_ss'
, 'disulfide_rr', 'disulfide_mm', 'disulfide_sm', 'disulfide_ss'
, 'hbonds_rr', 'hbonds_mm', 'hbonds_sm', 'hbonds_ss'
, 'partcov_rr', 'partcov_mm', 'partcov_sm', 'partcov_ss'
, 'vdwclashes_rr', 'vdwclashes_mm', 'vdwclashes_sm', 'vdwclashes_ss'
, 'volumetric_rr', 'volumetric_mm', 'volumetric_ss'
]
X_str = ['rsa'
#, 'asa'
, 'kd_values'
, 'rd_values']
X_ssFN = X_stabilityN + X_str + X_foldX_cols
X_evolFN = ['consurf_score'
, 'snap2_score'
, 'provean_score']
X_genomic_mafor = ['maf'
, 'logorI'
# , 'or_rawI'
# , 'or_mychisq'
# , 'or_logistic'
# , 'or_fisher'
# , 'pval_fisher'
]
X_genomic_linegae = ['lineage_proportion'
, 'dist_lineage_proportion'
#, 'lineage' # could be included as a category but it has L2;L4 formatting
, 'lineage_count_all'
, 'lineage_count_unique'
]
X_genomicFN = X_genomic_mafor + X_genomic_linegae
#X_aaindexFN = list(aa_df_cols)
#print('\nTotal no. of features for aaindex:', len(X_aaindexFN))
# numerical feature names [NO aa_index]
numerical_FN = X_ssFN + X_evolFN + X_genomicFN
# categorical feature names
categorical_FN = ['ss_class'
# , 'wt_prop_water'
# , 'mut_prop_water'
# , 'wt_prop_polarity'
# , 'mut_prop_polarity'
# , 'wt_calcprop'
# , 'mut_calcprop'
, 'aa_prop_change'
, 'electrostatics_change'
, 'polarity_change'
, 'water_change'
, 'drtype_mode_labels' # beware then you can't use it to predict [USED it for uq_v1, not v2]
, 'active_site' #[didn't use it for uq_v1]
#, 'gene_name' # will be required for the combined stuff
]
#----------------------------------------------
# count numerical and categorical features
#----------------------------------------------
print('\nNo. of numerical features:', len(numerical_FN)
, '\nNo. of categorical features:', len(categorical_FN))
###########################################################################
#=======================
# Masking columns:
# (mCSM-lig, mCSM-NA, mCSM-ppi2) values for lig_dist >10
#=======================
# my_df_ml['mutationinformation'][my_df['ligand_distance']>10].value_counts()
# my_df_ml.groupby('mutationinformation')['ligand_distance'].apply(lambda x: (x>10)).value_counts()
# my_df_ml.loc[(my_df_ml['ligand_distance'] > 10), 'ligand_affinity_change'] = 0
# (my_df_ml['ligand_affinity_change'] == 0).sum()
my_df_ml['mutationinformation'][my_df_ml['ligand_distance']>10].value_counts()
my_df_ml.groupby('mutationinformation')['ligand_distance'].apply(lambda x: (x>10)).value_counts()
my_df_ml.loc[(my_df_ml['ligand_distance'] > 10), cols_to_mask].value_counts()
# mask the mcsm affinity related columns where ligand distance > 10
my_df_ml.loc[(my_df_ml['ligand_distance'] > 10), cols_to_mask] = 0
(my_df_ml['ligand_affinity_change'] == 0).sum()
mask_check = my_df_ml[['mutationinformation', 'ligand_distance'] + cols_to_mask]
# write file for check
mask_check.sort_values(by = ['ligand_distance'], ascending = True, inplace = True)
mask_check.to_csv(outdir + 'ml/' + gene.lower() + '_mask_check.csv')
#===================================================
# Training and BLIND test set [UQ]: actual vs imputed
# No aa index but active_site included
# dst with actual values : training set
# dst with imputed values : blind test
#==================================================
my_df_ml[drug].isna().sum() #'na' ones are the blind_test set
blind_test_df = my_df_ml[my_df_ml[drug].isna()]
blind_test_df.shape
training_df = my_df_ml[my_df_ml[drug].notna()]
training_df.shape
# Target 1: dst_mode
training_df[drug].value_counts()
training_df['dst_mode'].value_counts()
####################################################################
#%% extracting dfs based on numerical, categorical column names
#----------------------------------
# WITHOUT the target var included
#----------------------------------
num_df = training_df[numerical_FN]
num_df.shape
cat_df = training_df[categorical_FN]
cat_df.shape
all_df = training_df[numerical_FN + categorical_FN]
all_df.shape
#------------------------------
# WITH the target var included:
#'wtgt': with target
#------------------------------
# drug and dst_mode should be the same thing
num_df_wtgt = training_df[numerical_FN + ['dst_mode']]
num_df_wtgt.shape
cat_df_wtgt = training_df[categorical_FN + ['dst_mode']]
cat_df_wtgt.shape
all_df_wtgt = training_df[numerical_FN + categorical_FN + ['dst_mode']]
all_df_wtgt.shape
#%%########################################################################
#============
# ML data
#============
#------
# X: Training and Blind test (BTS)
#------
X = all_df_wtgt[numerical_FN + categorical_FN] # training data ALL
X_bts = blind_test_df[numerical_FN + categorical_FN] # blind test data ALL
#X = all_df_wtgt[numerical_FN] # training numerical only
#X_bts = blind_test_df[numerical_FN] # blind test data numerical
#------
# y
#------
y = all_df_wtgt['dst_mode'] # training data y
y_bts = blind_test_df['dst_mode'] # blind data test y
#X_bts_wt = blind_test_df[numerical_FN + ['dst_mode']]
# Quick check
#(X['ligand_affinity_change']==0).sum() == (X['ligand_distance']>10).sum()
for i in range(len(cols_to_mask)):
ind = i+1
print('\nindex:', i, '\nind:', ind)
print('\nMask count check:'
, (my_df_ml[cols_to_mask[i]]==0).sum() == (my_df_ml['ligand_distance']>10).sum()
)
print('Original Data\n', Counter(y)
, 'Data dim:', X.shape)
yc1 = Counter(y)
yc1_ratio = yc1[0]/yc1[1]
yc2 = Counter(y_bts)
yc2_ratio = yc2[0]/yc2[1]
print('\n-------------------------------------------------------------'
, '\nSuccessfully split data: UQ [no aa_index but active site included] training'
, '\nactual values: training set'
, '\nimputed values: blind test set'
, '\nTrain data size:', X.shape
, '\nTest data size:', X_bts.shape
, '\ny_train numbers:', yc1
, '\ny_train ratio:',yc1_ratio
, '\n'
, '\ny_test_numbers:', yc2
, '\ny_test ratio:', yc2_ratio
, '\n-------------------------------------------------------------'
)
###########################################################################
#%%
###########################################################################
# RESAMPLING
###########################################################################
#------------------------------
# Simple Random oversampling
# [Numerical + catgeorical]
#------------------------------
oversample = RandomOverSampler(sampling_strategy='minority')
X_ros, y_ros = oversample.fit_resample(X, y)
print('Simple Random OverSampling\n', Counter(y_ros))
print(X_ros.shape)
#------------------------------
# Simple Random Undersampling
# [Numerical + catgeorical]
#------------------------------
undersample = RandomUnderSampler(sampling_strategy='majority')
X_rus, y_rus = undersample.fit_resample(X, y)
print('Simple Random UnderSampling\n', Counter(y_rus))
print(X_rus.shape)
#------------------------------
# Simple combine ROS and RUS
# [Numerical + catgeorical]
#------------------------------
oversample = RandomOverSampler(sampling_strategy='minority')
X_ros, y_ros = oversample.fit_resample(X, y)
undersample = RandomUnderSampler(sampling_strategy='majority')
X_rouC, y_rouC = undersample.fit_resample(X_ros, y_ros)
print('Simple Combined Over and UnderSampling\n', Counter(y_rouC))
print(X_rouC.shape)
#------------------------------
# SMOTE_NC: oversampling
# [numerical + categorical]
#https://stackoverflow.com/questions/47655813/oversampling-smote-for-binary-and-categorical-data-in-python
#------------------------------
# Determine categorical and numerical features
numerical_ix = X.select_dtypes(include=['int64', 'float64']).columns
numerical_ix
num_featuresL = list(numerical_ix)
numerical_colind = X.columns.get_indexer(list(numerical_ix) )
numerical_colind
categorical_ix = X.select_dtypes(include=['object', 'bool']).columns
categorical_ix
categorical_colind = X.columns.get_indexer(list(categorical_ix))
categorical_colind
k_sm = 5 # 5 is deafult
sm_nc = SMOTENC(categorical_features=categorical_colind, k_neighbors = k_sm, **rs, **njobs)
X_smnc, y_smnc = sm_nc.fit_resample(X, y)
print('SMOTE_NC OverSampling\n', Counter(y_smnc))
print(X_smnc.shape)
globals().update(locals()) # TROLOLOLOLOLOLS
#print("i did a horrible hack :-)")
###############################################################################
#%% SMOTE RESAMPLING for NUMERICAL ONLY*
# #------------------------------
# # SMOTE: Oversampling
# # [Numerical ONLY]
# #------------------------------
# k_sm = 1
# sm = SMOTE(sampling_strategy = 'auto', k_neighbors = k_sm, **rs)
# X_sm, y_sm = sm.fit_resample(X, y)
# print(X_sm.shape)
# print('SMOTE OverSampling\n', Counter(y_sm))
# y_sm_df = y_sm.to_frame()
# y_sm_df.value_counts().plot(kind = 'bar')
# #------------------------------
# # SMOTE: Over + Undersampling COMBINED
# # [Numerical ONLY]
# #-----------------------------
# sm_enn = SMOTEENN(enn=EditedNearestNeighbours(sampling_strategy='all', **rs, **njobs ))
# X_enn, y_enn = sm_enn.fit_resample(X, y)
# print(X_enn.shape)
# print('SMOTE Over+Under Sampling combined\n', Counter(y_enn))
###############################################################################
# TODO: Find over and undersampling JUST for categorical data

806
scripts/ml/ml_data_7030.py
View file

@ -1,806 +0,0 @@
#!/usr/bin/env python3
# -*- coding: utf-8 -*-
"""
Created on Sun Mar 6 13:41:54 2022
@author: tanu
"""
def setvars(gene,drug):
#https://stackoverflow.com/questions/51695322/compare-multiple-algorithms-with-sklearn-pipeline
import os, sys
import pandas as pd
import numpy as np
print(np.__version__)
print(pd.__version__)
import pprint as pp
from copy import deepcopy
from collections import Counter
from sklearn.impute import KNNImputer as KNN
from imblearn.over_sampling import RandomOverSampler
from imblearn.under_sampling import RandomUnderSampler
from imblearn.over_sampling import SMOTE
from sklearn.datasets import make_classification
from imblearn.combine import SMOTEENN
from imblearn.combine import SMOTETomek
from imblearn.over_sampling import SMOTENC
from imblearn.under_sampling import EditedNearestNeighbours
from imblearn.under_sampling import RepeatedEditedNearestNeighbours
from sklearn.metrics import make_scorer, confusion_matrix, accuracy_score, balanced_accuracy_score, precision_score, average_precision_score, recall_score
from sklearn.metrics import roc_auc_score, roc_curve, f1_score, matthews_corrcoef, jaccard_score, classification_report
from sklearn.model_selection import train_test_split, cross_validate, cross_val_score
from sklearn.model_selection import StratifiedKFold,RepeatedStratifiedKFold, RepeatedKFold
from sklearn.pipeline import Pipeline, make_pipeline
import argparse
import re
#%% GLOBALS
tts_split = "70_30"
rs = {'random_state': 42}
njobs = {'n_jobs': 10}
scoring_fn = ({ 'mcc' : make_scorer(matthews_corrcoef)
, 'accuracy' : make_scorer(accuracy_score)
, 'fscore' : make_scorer(f1_score)
, 'precision' : make_scorer(precision_score)
, 'recall' : make_scorer(recall_score)
, 'roc_auc' : make_scorer(roc_auc_score)
, 'jcc' : make_scorer(jaccard_score)
})
skf_cv = StratifiedKFold(n_splits = 10
#, shuffle = False, random_state= None)
, shuffle = True,**rs)
rskf_cv = RepeatedStratifiedKFold(n_splits = 10
, n_repeats = 3
, **rs)
mcc_score_fn = {'mcc': make_scorer(matthews_corrcoef)}
jacc_score_fn = {'jcc': make_scorer(jaccard_score)}
#%% FOR LATER: Combine ED logo data
###########################################################################
homedir = os.path.expanduser("~")
geneL_basic = ['pnca']
geneL_na = ['gid']
geneL_na_ppi2 = ['rpob']
geneL_ppi2 = ['alr', 'embb', 'katg']
#num_type = ['int64', 'float64']
num_type = ['int16', 'int32', 'int64', 'float16', 'float32', 'float64']
cat_type = ['object', 'bool']
#==============
# directories
#==============
datadir = homedir + '/git/Data/'
indir = datadir + drug + '/input/'
outdir = datadir + drug + '/output/'
#=======
# input
#=======
#---------
# File 1
#---------
infile_ml1 = outdir + gene.lower() + '_merged_df3.csv'
#infile_ml2 = outdir + gene.lower() + '_merged_df2.csv'
my_features_df = pd.read_csv(infile_ml1, index_col = 0)
my_features_df = my_features_df .reset_index(drop = True)
my_features_df.index
my_features_df.dtypes
mycols = my_features_df.columns
#---------
# File 2
#---------
infile_aaindex = outdir + 'aa_index/' + gene.lower() + '_aa.csv'
aaindex_df = pd.read_csv(infile_aaindex, index_col = 0)
aaindex_df.dtypes
#-----------
# check for non-numerical columns
#-----------
if any(aaindex_df.dtypes==object):
print('\naaindex_df contains non-numerical data')
aaindex_df_object = aaindex_df.select_dtypes(include = cat_type)
print('\nTotal no. of non-numerial columns:', len(aaindex_df_object.columns))
expected_aa_ncols = len(aaindex_df.columns) - len(aaindex_df_object.columns)
#-----------
# Extract numerical data only
#-----------
print('\nSelecting numerical data only')
aaindex_df = aaindex_df.select_dtypes(include = num_type)
#---------------------------
# aaindex: sanity check 1
#---------------------------
if len(aaindex_df.columns) == expected_aa_ncols:
print('\nPASS: successfully selected numerical columns only for aaindex_df')
else:
print('\nFAIL: Numbers mismatch'
, '\nExpected ncols:', expected_aa_ncols
, '\nGot:', len(aaindex_df.columns))
#---------------
# check for NA
#---------------
print('\nNow checking for NA in the remaining aaindex_cols')
c1 = aaindex_df.isna().sum()
c2 = c1.sort_values(ascending=False)
print('\nCounting aaindex_df cols with NA'
, '\nncols with NA:', sum(c2>0), 'columns'
, '\nDropping these...'
, '\nOriginal ncols:', len(aaindex_df.columns)
)
aa_df = aaindex_df.dropna(axis=1)
print('\nRevised df ncols:', len(aa_df.columns))
c3 = aa_df.isna().sum()
c4 = c3.sort_values(ascending=False)
print('\nChecking NA in revised df...')
if sum(c4>0):
sys.exit('\nFAIL: aaindex_df still contains cols with NA, please check and drop these before proceeding...')
else:
print('\nPASS: cols with NA successfully dropped from aaindex_df'
, '\nProceeding with combining aa_df with other features_df')
#---------------------------
# aaindex: sanity check 2
#---------------------------
expected_aa_ncols2 = len(aaindex_df.columns) - sum(c2>0)
if len(aa_df.columns) == expected_aa_ncols2:
print('\nPASS: ncols match'
, '\nExpected ncols:', expected_aa_ncols2
, '\nGot:', len(aa_df.columns))
else:
print('\nFAIL: Numbers mismatch'
, '\nExpected ncols:', expected_aa_ncols2
, '\nGot:', len(aa_df.columns))
# Important: need this to identify aaindex cols
aa_df_cols = aa_df.columns
print('\nTotal no. of columns in clean aa_df:', len(aa_df_cols))
###############################################################################
#%% Combining my_features_df and aaindex_df
#===========================
# Merge my_df + aaindex_df
#===========================
if aa_df.columns[aa_df.columns.isin(my_features_df.columns)] == my_features_df.columns[my_features_df.columns.isin(aa_df.columns)]:
print('\nMerging on column: mutationinformation')
if len(my_features_df) == len(aa_df):
expected_nrows = len(my_features_df)
print('\nProceeding to merge, expected nrows in merged_df:', expected_nrows)
else:
sys.exit('\nNrows mismatch, cannot merge. Please check'
, '\nnrows my_df:', len(my_features_df)
, '\nnrows aa_df:', len(aa_df))
#-----------------
# Reset index: mutationinformation
# Very important for merging
#-----------------
aa_df = aa_df.reset_index()
expected_ncols = len(my_features_df.columns) + len(aa_df.columns) - 1 # for the no. of merging col
#-----------------
# Merge: my_features_df + aa_df
#-----------------
merged_df = pd.merge(my_features_df
, aa_df
, on = 'mutationinformation')
#---------------------------
# aaindex: sanity check 3
#---------------------------
if len(merged_df.columns) == expected_ncols:
print('\nPASS: my_features_df and aa_df successfully combined'
, '\nnrows:', len(merged_df)
, '\nncols:', len(merged_df.columns))
else:
sys.exit('\nFAIL: could not combine my_features_df and aa_df'
, '\nCheck dims and merging cols!')
#--------
# Reassign so downstream code doesn't need to change
#--------
my_df = merged_df.copy()
#%% Data: my_df
# Check if non structural pos have crept in
# IDEALLY remove from source! But for rpoB do it here
# Drop NA where numerical cols have them
if gene.lower() in geneL_na_ppi2:
#D1148 get rid of
na_index = my_df['mutationinformation'].index[my_df['mcsm_na_affinity'].apply(np.isnan)]
my_df = my_df.drop(index=na_index)
# FIXED: complete data for all muts inc L114M, F115L, V123L, V125I, V131M
# if gene.lower() in ['embb']:
# na_index = my_df['mutationinformation'].index[my_df['ligand_distance'].apply(np.isnan)]
# my_df = my_df.drop(index=na_index)
# # Sanity check for non-structural positions
# print('\nChecking for non-structural postions')
# na_index = my_df['mutationinformation'].index[my_df['ligand_distance'].apply(np.isnan)]
# if len(na_index) > 0:
# print('\nNon-structural positions detected for gene:', gene.lower()
# , '\nTotal number of these detected:', len(na_index)
# , '\These are at index:', na_index
# , '\nOriginal nrows:', len(my_df)
# , '\nDropping these...')
# my_df = my_df.drop(index=na_index)
# print('\nRevised nrows:', len(my_df))
# else:
# print('\nNo non-structural positions detected for gene:', gene.lower()
# , '\nnrows:', len(my_df))
###########################################################################
#%% Add lineage calculation columns
#FIXME: Check if this can be imported from config?
total_mtblineage_uc = 8
lineage_colnames = ['lineage_list_all', 'lineage_count_all', 'lineage_count_unique', 'lineage_list_unique', 'lineage_multimode']
#bar = my_df[lineage_colnames]
my_df['lineage_proportion'] = my_df['lineage_count_unique']/my_df['lineage_count_all']
my_df['dist_lineage_proportion'] = my_df['lineage_count_unique']/total_mtblineage_uc
###########################################################################
#%% Active site annotation column
# change from numberic to categorical
if my_df['active_site'].dtype in num_type:
my_df['active_site'] = my_df['active_site'].astype(object)
my_df['active_site'].dtype
#%% AA property change
#--------------------
# Water prop change
#--------------------
my_df['water_change'] = my_df['wt_prop_water'] + str('_to_') + my_df['mut_prop_water']
my_df['water_change'].value_counts()
water_prop_changeD = {
'hydrophobic_to_neutral' : 'change'
, 'hydrophobic_to_hydrophobic' : 'no_change'
, 'neutral_to_neutral' : 'no_change'
, 'neutral_to_hydrophobic' : 'change'
, 'hydrophobic_to_hydrophilic' : 'change'
, 'neutral_to_hydrophilic' : 'change'
, 'hydrophilic_to_neutral' : 'change'
, 'hydrophilic_to_hydrophobic' : 'change'
, 'hydrophilic_to_hydrophilic' : 'no_change'
}
my_df['water_change'] = my_df['water_change'].map(water_prop_changeD)
my_df['water_change'].value_counts()
#--------------------
# Polarity change
#--------------------
my_df['polarity_change'] = my_df['wt_prop_polarity'] + str('_to_') + my_df['mut_prop_polarity']
my_df['polarity_change'].value_counts()
polarity_prop_changeD = {
'non-polar_to_non-polar' : 'no_change'
, 'non-polar_to_neutral' : 'change'
, 'neutral_to_non-polar' : 'change'
, 'neutral_to_neutral' : 'no_change'
, 'non-polar_to_basic' : 'change'
, 'acidic_to_neutral' : 'change'
, 'basic_to_neutral' : 'change'
, 'non-polar_to_acidic' : 'change'
, 'neutral_to_basic' : 'change'
, 'acidic_to_non-polar' : 'change'
, 'basic_to_non-polar' : 'change'
, 'neutral_to_acidic' : 'change'
, 'acidic_to_acidic' : 'no_change'
, 'basic_to_acidic' : 'change'
, 'basic_to_basic' : 'no_change'
, 'acidic_to_basic' : 'change'}
my_df['polarity_change'] = my_df['polarity_change'].map(polarity_prop_changeD)
my_df['polarity_change'].value_counts()
#--------------------
# Electrostatics change
#--------------------
my_df['electrostatics_change'] = my_df['wt_calcprop'] + str('_to_') + my_df['mut_calcprop']
my_df['electrostatics_change'].value_counts()
calc_prop_changeD = {
'non-polar_to_non-polar' : 'no_change'
, 'non-polar_to_polar' : 'change'
, 'polar_to_non-polar' : 'change'
, 'non-polar_to_pos' : 'change'
, 'neg_to_non-polar' : 'change'
, 'non-polar_to_neg' : 'change'
, 'pos_to_polar' : 'change'
, 'pos_to_non-polar' : 'change'
, 'polar_to_polar' : 'no_change'
, 'neg_to_neg' : 'no_change'
, 'polar_to_neg' : 'change'
, 'pos_to_neg' : 'change'
, 'pos_to_pos' : 'no_change'
, 'polar_to_pos' : 'change'
, 'neg_to_polar' : 'change'
, 'neg_to_pos' : 'change'
}
my_df['electrostatics_change'] = my_df['electrostatics_change'].map(calc_prop_changeD)
my_df['electrostatics_change'].value_counts()
#--------------------
# Summary change: Create a combined column summarising these three cols
#--------------------
detect_change = 'change'
check_prop_cols = ['water_change', 'polarity_change', 'electrostatics_change']
#my_df['aa_prop_change'] = (my_df.values == detect_change).any(1).astype(int)
my_df['aa_prop_change'] = (my_df[check_prop_cols].values == detect_change).any(1).astype(int)
my_df['aa_prop_change'].value_counts()
my_df['aa_prop_change'].dtype
my_df['aa_prop_change'] = my_df['aa_prop_change'].map({1:'change'
, 0: 'no_change'})
my_df['aa_prop_change'].value_counts()
my_df['aa_prop_change'].dtype
#%% IMPUTE values for OR [check script for exploration: UQ_or_imputer]
#--------------------
# Impute OR values
#--------------------
#or_cols = ['or_mychisq', 'log10_or_mychisq', 'or_fisher']
sel_cols = ['mutationinformation', 'or_mychisq', 'log10_or_mychisq']
or_cols = ['or_mychisq', 'log10_or_mychisq']
print("count of NULL values before imputation\n")
print(my_df[or_cols].isnull().sum())
my_dfI = pd.DataFrame(index = my_df['mutationinformation'] )
my_dfI = pd.DataFrame(KNN(n_neighbors=3, weights="uniform").fit_transform(my_df[or_cols])
, index = my_df['mutationinformation']
, columns = or_cols )
my_dfI.columns = ['or_rawI', 'logorI']
my_dfI.columns
my_dfI = my_dfI.reset_index(drop = False) # prevents old index from being added as a column
my_dfI.head()
print("count of NULL values AFTER imputation\n")
print(my_dfI.isnull().sum())
#-------------------------------------------
# OR df Merge: with original based on index
#-------------------------------------------
#my_df['index_bm'] = my_df.index
mydf_imputed = pd.merge(my_df
, my_dfI
, on = 'mutationinformation')
#mydf_imputed = mydf_imputed.set_index(['index_bm'])
my_df['log10_or_mychisq'].isna().sum()
mydf_imputed['log10_or_mychisq'].isna().sum()
mydf_imputed['logorI'].isna().sum() # should be 0
len(my_df.columns)
len(mydf_imputed.columns)
#-----------------------------------------
# REASSIGN my_df after imputing OR values
#-----------------------------------------
my_df = mydf_imputed.copy()
if my_df['logorI'].isna().sum() == 0:
print('\nPASS: OR values imputed, data ready for ML')
else:
sys.exit('\nFAIL: something went wrong, Data not ready for ML. Please check upstream!')
#!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
#---------------------------------------
# TODO: try other imputation like MICE
#---------------------------------------
#!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
#%%########################################################################
#==========================
# Data for ML
#==========================
my_df_ml = my_df.copy()
# Build column names to mask for affinity chanhes
if gene.lower() in geneL_basic:
#X_stabilityN = common_cols_stabiltyN
gene_affinity_colnames = []# not needed as its the common ones
cols_to_mask = ['ligand_affinity_change']
if gene.lower() in geneL_ppi2:
gene_affinity_colnames = ['mcsm_ppi2_affinity', 'interface_dist']
#X_stabilityN = common_cols_stabiltyN + geneL_ppi2_st_cols
cols_to_mask = ['ligand_affinity_change', 'mcsm_ppi2_affinity']
if gene.lower() in geneL_na:
gene_affinity_colnames = ['mcsm_na_affinity']
#X_stabilityN = common_cols_stabiltyN + geneL_na_st_cols
cols_to_mask = ['ligand_affinity_change', 'mcsm_na_affinity']
if gene.lower() in geneL_na_ppi2:
gene_affinity_colnames = ['mcsm_na_affinity'] + ['mcsm_ppi2_affinity', 'interface_dist']
#X_stabilityN = common_cols_stabiltyN + geneL_na_ppi2_st_cols
cols_to_mask = ['ligand_affinity_change', 'mcsm_na_affinity', 'mcsm_ppi2_affinity']
#=======================
# Masking columns:
# (mCSM-lig, mCSM-NA, mCSM-ppi2) values for lig_dist >10
#=======================
my_df_ml['mutationinformation'][my_df_ml['ligand_distance']>10].value_counts()
my_df_ml.groupby('mutationinformation')['ligand_distance'].apply(lambda x: (x>10)).value_counts()
my_df_ml.loc[(my_df_ml['ligand_distance'] > 10), cols_to_mask].value_counts()
# mask the mcsm affinity related columns where ligand distance > 10
my_df_ml.loc[(my_df_ml['ligand_distance'] > 10), cols_to_mask] = 0
(my_df_ml['ligand_affinity_change'] == 0).sum()
mask_check = my_df_ml[['mutationinformation', 'ligand_distance'] + cols_to_mask]
#===================================================
# write file for check
mask_check.sort_values(by = ['ligand_distance'], ascending = True, inplace = True)
mask_check.to_csv(outdir + 'ml/' + gene.lower() + '_mask_check.csv')
#===================================================
###############################################################################
#%% Feature groups (FG): Build X for Input ML
############################################################################
#===========================
# FG1: Evolutionary features
#===========================
X_evolFN = ['consurf_score'
, 'snap2_score'
, 'provean_score']
###############################################################################
#========================
# FG2: Stability features
#========================
#--------
# common
#--------
X_common_stability_Fnum = [
'duet_stability_change'
, 'ddg_foldx'
, 'deepddg'
, 'ddg_dynamut2'
, 'contacts']
#--------
# FoldX
#--------
X_foldX_Fnum = [ 'electro_rr', 'electro_mm', 'electro_sm', 'electro_ss'
, 'disulfide_rr', 'disulfide_mm', 'disulfide_sm', 'disulfide_ss'
, 'hbonds_rr', 'hbonds_mm', 'hbonds_sm', 'hbonds_ss'
, 'partcov_rr', 'partcov_mm', 'partcov_sm', 'partcov_ss'
, 'vdwclashes_rr', 'vdwclashes_mm', 'vdwclashes_sm', 'vdwclashes_ss'
, 'volumetric_rr', 'volumetric_mm', 'volumetric_ss']
X_stability_FN = X_common_stability_Fnum + X_foldX_Fnum
###############################################################################
#===================
# FG3: Affinity features
#===================
common_affinity_Fnum = ['ligand_distance'
, 'ligand_affinity_change'
, 'mmcsm_lig']
# if gene.lower() in geneL_basic:
# X_affinityFN = common_affinity_Fnum
# else:
# X_affinityFN = common_affinity_Fnum + gene_affinity_colnames
X_affinityFN = common_affinity_Fnum + gene_affinity_colnames
###############################################################################
#============================
# FG4: Residue level features
#============================
#-----------
# AA index
#-----------
X_aaindex_Fnum = list(aa_df_cols)
print('\nTotal no. of features for aaindex:', len(X_aaindex_Fnum))
#-----------------
# surface area
# depth
# hydrophobicity
#-----------------
X_str_Fnum = ['rsa'
#, 'asa'
, 'kd_values'
, 'rd_values']
#---------------------------
# Other aa properties
# active site indication
#---------------------------
X_aap_Fcat = ['ss_class'
# , 'wt_prop_water'
# , 'mut_prop_water'
# , 'wt_prop_polarity'
# , 'mut_prop_polarity'
# , 'wt_calcprop'
# , 'mut_calcprop'
, 'aa_prop_change'
, 'electrostatics_change'
, 'polarity_change'
, 'water_change'
, 'active_site']
X_resprop_FN = X_aaindex_Fnum + X_str_Fnum + X_aap_Fcat
###############################################################################
#========================
# FG5: Genomic features
#========================
X_gn_mafor_Fnum = ['maf'
#, 'logorI'
# , 'or_rawI'
# , 'or_mychisq'
# , 'or_logistic'
# , 'or_fisher'
# , 'pval_fisher'
]
X_gn_linegae_Fnum = ['lineage_proportion'
, 'dist_lineage_proportion'
#, 'lineage' # could be included as a category but it has L2;L4 formatting
, 'lineage_count_all'
, 'lineage_count_unique'
]
# X_gn_Fcat = ['drtype_mode_labels' # beware then you can't use it to predict [USED it for uq_v1, not v2]
# #, 'gene_name' # will be required for the combined stuff
# ]
X_gn_Fcat = []
X_genomicFN = X_gn_mafor_Fnum + X_gn_linegae_Fnum + X_gn_Fcat
###############################################################################
#========================
# FG6 collapsed: Structural : Atability + Affinity + ResidueProp
#========================
X_structural_FN = X_stability_FN + X_affinityFN + X_resprop_FN
###############################################################################
#========================
# BUILDING all features
#========================
all_featuresN = X_evolFN + X_structural_FN + X_genomicFN
###############################################################################
#%% Define training and test data
#================================================================
# Training and BLIND test set: 70/30
# dst with actual values : training set
# dst with imputed values : THROW AWAY [unrepresentative]
#================================================================
my_df_ml[drug].isna().sum()
# blind_test_df = my_df_ml[my_df_ml[drug].isna()]
# blind_test_df.shape
training_df = my_df_ml[my_df_ml[drug].notna()]
training_df.shape
# Target 1: dst_mode
training_df[drug].value_counts()
training_df['dst_mode'].value_counts()
####################################################################
#====================================
# ML data: Train test split: 70/30
# with stratification
# 70% : training_data for CV
# 30% : blind test
#=====================================
x_features = training_df[all_featuresN]
y_target = training_df['dst_mode']
# sanity check
if not 'dst_mode' in x_features.columns:
print('\nPASS: x_features has no target variable')
x_ncols = len(x_features.columns)
print('\nNo. of columns for x_features:', x_ncols)
# NEED It for scaling law split
#https://towardsdatascience.com/finally-why-we-use-an-80-20-split-for-training-and-test-data-plus-an-alternative-method-oh-yes-edc77e96295d
else:
sys.exit('\nFAIL: x_features has target variable included. FIX it and rerun!')
#-------------------
# train-test split
#-------------------
#x_train, x_test, y_train, y_test # traditional var_names
# so my downstream code doesn't need to change
X, X_bts, y, y_bts = train_test_split(x_features, y_target
, test_size = 0.33
, **rs
, stratify = y_target)
yc1 = Counter(y)
yc1_ratio = yc1[0]/yc1[1]
yc2 = Counter(y_bts)
yc2_ratio = yc2[0]/yc2[1]
###############################################################################
#======================================================
# Determine categorical and numerical features
#======================================================
numerical_cols = X.select_dtypes(include=['int64', 'float64']).columns
numerical_cols
categorical_cols = X.select_dtypes(include=['object', 'bool']).columns
categorical_cols
################################################################################
# IMPORTANT sanity checks
if len(X.columns) == len(X_evolFN) + len(X_stability_FN) + len(X_affinityFN) + len(X_resprop_FN) + len(X_genomicFN):
print('\nPASS: ML data with input features, training and test generated...'
, '\n\nTotal no. of input features:' , len(X.columns)
, '\n--------No. of numerical features:' , len(numerical_cols)
, '\n--------No. of categorical features:' , len(categorical_cols)
, '\n\nTotal no. of evolutionary features:' , len(X_evolFN)
, '\n\nTotal no. of stability features:' , len(X_stability_FN)
, '\n--------Common stabilty cols:' , len(X_common_stability_Fnum)
, '\n--------Foldx cols:' , len(X_foldX_Fnum)
, '\n\nTotal no. of affinity features:' , len(X_affinityFN)
, '\n--------Common affinity cols:' , len(common_affinity_Fnum)
, '\n--------Gene specific affinity cols:' , len(gene_affinity_colnames)
, '\n\nTotal no. of residue level features:', len(X_resprop_FN)
, '\n--------AA index cols:' , len(X_aaindex_Fnum)
, '\n--------Residue Prop cols:' , len(X_str_Fnum)
, '\n--------AA change Prop cols:' , len(X_aap_Fcat)
, '\n\nTotal no. of genomic features:' , len(X_genomicFN)
, '\n--------MAF+OR cols:' , len(X_gn_mafor_Fnum)
, '\n--------Lineage cols:' , len(X_gn_linegae_Fnum)
, '\n--------Other cols:' , len(X_gn_Fcat)
)
else:
print('\nFAIL: numbers mismatch'
, '\nExpected:',len(X_evolFN) + len(X_stability_FN) + len(X_affinityFN) + len(X_resprop_FN) + len(X_genomicFN)
, '\nGot:', len(X.columns))
sys.exit()
###############################################################################
print('\n-------------------------------------------------------------'
, '\nSuccessfully split data: ALL features'
, '\nactual values: training set'
, '\nSplit:', tts_split
#, '\nimputed values: blind test set'
, '\n\nTotal data size:', len(X) + len(X_bts)
, '\n\nTrain data size:', X.shape
, '\ny_train numbers:', yc1
, '\n\nTest data size:', X_bts.shape
, '\ny_test_numbers:', yc2
, '\n\ny_train ratio:',yc1_ratio
, '\ny_test ratio:', yc2_ratio
, '\n-------------------------------------------------------------'
)
##########################################################################
# Quick check
#(X['ligand_affinity_change']==0).sum() == (X['ligand_distance']>10).sum()
for i in range(len(cols_to_mask)):
ind = i+1
print('\nindex:', i, '\nind:', ind)
print('\nMask count check:'
, (my_df_ml[cols_to_mask[i]]==0).sum() == (my_df_ml['ligand_distance']>10).sum()
)
print('Original Data\n', Counter(y)
, 'Data dim:', X.shape)
###########################################################################
#%%
###########################################################################
# RESAMPLING
###########################################################################
#------------------------------
# Simple Random oversampling
# [Numerical + catgeorical]
#------------------------------
oversample = RandomOverSampler(sampling_strategy='minority')
X_ros, y_ros = oversample.fit_resample(X, y)
print('\nSimple Random OverSampling\n', Counter(y_ros))
print(X_ros.shape)
#------------------------------
# Simple Random Undersampling
# [Numerical + catgeorical]
#------------------------------
undersample = RandomUnderSampler(sampling_strategy='majority')
X_rus, y_rus = undersample.fit_resample(X, y)
print('\nSimple Random UnderSampling\n', Counter(y_rus))
print(X_rus.shape)
#------------------------------
# Simple combine ROS and RUS
# [Numerical + catgeorical]
#------------------------------
oversample = RandomOverSampler(sampling_strategy='minority')
X_ros, y_ros = oversample.fit_resample(X, y)
undersample = RandomUnderSampler(sampling_strategy='majority')
X_rouC, y_rouC = undersample.fit_resample(X_ros, y_ros)
print('\nSimple Combined Over and UnderSampling\n', Counter(y_rouC))
print(X_rouC.shape)
#------------------------------
# SMOTE_NC: oversampling
# [numerical + categorical]
#https://stackoverflow.com/questions/47655813/oversampling-smote-for-binary-and-categorical-data-in-python
#------------------------------
# Determine categorical and numerical features
numerical_ix = X.select_dtypes(include=['int64', 'float64']).columns
numerical_ix
num_featuresL = list(numerical_ix)
numerical_colind = X.columns.get_indexer(list(numerical_ix) )
numerical_colind
categorical_ix = X.select_dtypes(include=['object', 'bool']).columns
categorical_ix
categorical_colind = X.columns.get_indexer(list(categorical_ix))
categorical_colind
k_sm = 5 # 5 is default
sm_nc = SMOTENC(categorical_features=categorical_colind, k_neighbors = k_sm, **rs, **njobs)
X_smnc, y_smnc = sm_nc.fit_resample(X, y)
print('\nSMOTE_NC OverSampling\n', Counter(y_smnc))
print(X_smnc.shape)
globals().update(locals()) # TROLOLOLOLOLOLS
#print("i did a horrible hack :-)")
###############################################################################
#%% SMOTE RESAMPLING for NUMERICAL ONLY*
# #------------------------------
# # SMOTE: Oversampling
# # [Numerical ONLY]
# #------------------------------
# k_sm = 1
# sm = SMOTE(sampling_strategy = 'auto', k_neighbors = k_sm, **rs)
# X_sm, y_sm = sm.fit_resample(X, y)
# print(X_sm.shape)
# print('\nSMOTE OverSampling\n', Counter(y_sm))
# y_sm_df = y_sm.to_frame()
# y_sm_df.value_counts().plot(kind = 'bar')
# #------------------------------
# # SMOTE: Over + Undersampling COMBINED
# # [Numerical ONLY]
# #-----------------------------
# sm_enn = SMOTEENN(enn=EditedNearestNeighbours(sampling_strategy='all', **rs, **njobs ))
# X_enn, y_enn = sm_enn.fit_resample(X, y)
# print(X_enn.shape)
# print('\nSMOTE Over+Under Sampling combined\n', Counter(y_enn))
###########################################################################
# TODO: Find over and undersampling JUST for categorical data
###########################################################################
print('\n#################################################################'
, '\nDim of X for gene:', gene.lower(), '\n', X.shape
, '\n###############################################################')

806
scripts/ml/ml_data_8020.py
View file

@ -1,806 +0,0 @@
#!/usr/bin/env python3
# -*- coding: utf-8 -*-
"""
Created on Sun Mar 6 13:41:54 2022
@author: tanu
"""
def setvars(gene,drug):
#https://stackoverflow.com/questions/51695322/compare-multiple-algorithms-with-sklearn-pipeline
import os, sys
import pandas as pd
import numpy as np
print(np.__version__)
print(pd.__version__)
import pprint as pp
from copy import deepcopy
from collections import Counter
from sklearn.impute import KNNImputer as KNN
from imblearn.over_sampling import RandomOverSampler
from imblearn.under_sampling import RandomUnderSampler
from imblearn.over_sampling import SMOTE
from sklearn.datasets import make_classification
from imblearn.combine import SMOTEENN
from imblearn.combine import SMOTETomek
from imblearn.over_sampling import SMOTENC
from imblearn.under_sampling import EditedNearestNeighbours
from imblearn.under_sampling import RepeatedEditedNearestNeighbours
from sklearn.metrics import make_scorer, confusion_matrix, accuracy_score, balanced_accuracy_score, precision_score, average_precision_score, recall_score
from sklearn.metrics import roc_auc_score, roc_curve, f1_score, matthews_corrcoef, jaccard_score, classification_report
from sklearn.model_selection import train_test_split, cross_validate, cross_val_score
from sklearn.model_selection import StratifiedKFold,RepeatedStratifiedKFold, RepeatedKFold
from sklearn.pipeline import Pipeline, make_pipeline
import argparse
import re
#%% GLOBALS
tts_split = "80_20"
rs = {'random_state': 42}
njobs = {'n_jobs': 10}
scoring_fn = ({ 'mcc' : make_scorer(matthews_corrcoef)
, 'accuracy' : make_scorer(accuracy_score)
, 'fscore' : make_scorer(f1_score)
, 'precision' : make_scorer(precision_score)
, 'recall' : make_scorer(recall_score)
, 'roc_auc' : make_scorer(roc_auc_score)
, 'jcc' : make_scorer(jaccard_score)
})
skf_cv = StratifiedKFold(n_splits = 10
#, shuffle = False, random_state= None)
, shuffle = True,**rs)
rskf_cv = RepeatedStratifiedKFold(n_splits = 10
, n_repeats = 3
, **rs)
mcc_score_fn = {'mcc': make_scorer(matthews_corrcoef)}
jacc_score_fn = {'jcc': make_scorer(jaccard_score)}
#%% FOR LATER: Combine ED logo data
###########################################################################
homedir = os.path.expanduser("~")
geneL_basic = ['pnca']
geneL_na = ['gid']
geneL_na_ppi2 = ['rpob']
geneL_ppi2 = ['alr', 'embb', 'katg']
#num_type = ['int64', 'float64']
num_type = ['int16', 'int32', 'int64', 'float16', 'float32', 'float64']
cat_type = ['object', 'bool']
#==============
# directories
#==============
datadir = homedir + '/git/Data/'
indir = datadir + drug + '/input/'
outdir = datadir + drug + '/output/'
#=======
# input
#=======
#---------
# File 1
#---------
infile_ml1 = outdir + gene.lower() + '_merged_df3.csv'
#infile_ml2 = outdir + gene.lower() + '_merged_df2.csv'
my_features_df = pd.read_csv(infile_ml1, index_col = 0)
my_features_df = my_features_df .reset_index(drop = True)
my_features_df.index
my_features_df.dtypes
mycols = my_features_df.columns
#---------
# File 2
#---------
infile_aaindex = outdir + 'aa_index/' + gene.lower() + '_aa.csv'
aaindex_df = pd.read_csv(infile_aaindex, index_col = 0)
aaindex_df.dtypes
#-----------
# check for non-numerical columns
#-----------
if any(aaindex_df.dtypes==object):
print('\naaindex_df contains non-numerical data')
aaindex_df_object = aaindex_df.select_dtypes(include = cat_type)
print('\nTotal no. of non-numerial columns:', len(aaindex_df_object.columns))
expected_aa_ncols = len(aaindex_df.columns) - len(aaindex_df_object.columns)
#-----------
# Extract numerical data only
#-----------
print('\nSelecting numerical data only')
aaindex_df = aaindex_df.select_dtypes(include = num_type)
#---------------------------
# aaindex: sanity check 1
#---------------------------
if len(aaindex_df.columns) == expected_aa_ncols:
print('\nPASS: successfully selected numerical columns only for aaindex_df')
else:
print('\nFAIL: Numbers mismatch'
, '\nExpected ncols:', expected_aa_ncols
, '\nGot:', len(aaindex_df.columns))
#---------------
# check for NA
#---------------
print('\nNow checking for NA in the remaining aaindex_cols')
c1 = aaindex_df.isna().sum()
c2 = c1.sort_values(ascending=False)
print('\nCounting aaindex_df cols with NA'
, '\nncols with NA:', sum(c2>0), 'columns'
, '\nDropping these...'
, '\nOriginal ncols:', len(aaindex_df.columns)
)
aa_df = aaindex_df.dropna(axis=1)
print('\nRevised df ncols:', len(aa_df.columns))
c3 = aa_df.isna().sum()
c4 = c3.sort_values(ascending=False)
print('\nChecking NA in revised df...')
if sum(c4>0):
sys.exit('\nFAIL: aaindex_df still contains cols with NA, please check and drop these before proceeding...')
else:
print('\nPASS: cols with NA successfully dropped from aaindex_df'
, '\nProceeding with combining aa_df with other features_df')
#---------------------------
# aaindex: sanity check 2
#---------------------------
expected_aa_ncols2 = len(aaindex_df.columns) - sum(c2>0)
if len(aa_df.columns) == expected_aa_ncols2:
print('\nPASS: ncols match'
, '\nExpected ncols:', expected_aa_ncols2
, '\nGot:', len(aa_df.columns))
else:
print('\nFAIL: Numbers mismatch'
, '\nExpected ncols:', expected_aa_ncols2
, '\nGot:', len(aa_df.columns))
# Important: need this to identify aaindex cols
aa_df_cols = aa_df.columns
print('\nTotal no. of columns in clean aa_df:', len(aa_df_cols))
###############################################################################
#%% Combining my_features_df and aaindex_df
#===========================
# Merge my_df + aaindex_df
#===========================
if aa_df.columns[aa_df.columns.isin(my_features_df.columns)] == my_features_df.columns[my_features_df.columns.isin(aa_df.columns)]:
print('\nMerging on column: mutationinformation')
if len(my_features_df) == len(aa_df):
expected_nrows = len(my_features_df)
print('\nProceeding to merge, expected nrows in merged_df:', expected_nrows)
else:
sys.exit('\nNrows mismatch, cannot merge. Please check'
, '\nnrows my_df:', len(my_features_df)
, '\nnrows aa_df:', len(aa_df))
#-----------------
# Reset index: mutationinformation
# Very important for merging
#-----------------
aa_df = aa_df.reset_index()
expected_ncols = len(my_features_df.columns) + len(aa_df.columns) - 1 # for the no. of merging col
#-----------------
# Merge: my_features_df + aa_df
#-----------------
merged_df = pd.merge(my_features_df
, aa_df
, on = 'mutationinformation')
#---------------------------
# aaindex: sanity check 3
#---------------------------
if len(merged_df.columns) == expected_ncols:
print('\nPASS: my_features_df and aa_df successfully combined'
, '\nnrows:', len(merged_df)
, '\nncols:', len(merged_df.columns))
else:
sys.exit('\nFAIL: could not combine my_features_df and aa_df'
, '\nCheck dims and merging cols!')
#--------
# Reassign so downstream code doesn't need to change
#--------
my_df = merged_df.copy()
#%% Data: my_df
# Check if non structural pos have crept in
# IDEALLY remove from source! But for rpoB do it here
# Drop NA where numerical cols have them
if gene.lower() in geneL_na_ppi2:
#D1148 get rid of
na_index = my_df['mutationinformation'].index[my_df['mcsm_na_affinity'].apply(np.isnan)]
my_df = my_df.drop(index=na_index)
# FIXED: complete data for all muts inc L114M, F115L, V123L, V125I, V131M
# if gene.lower() in ['embb']:
# na_index = my_df['mutationinformation'].index[my_df['ligand_distance'].apply(np.isnan)]
# my_df = my_df.drop(index=na_index)
# # Sanity check for non-structural positions
# print('\nChecking for non-structural postions')
# na_index = my_df['mutationinformation'].index[my_df['ligand_distance'].apply(np.isnan)]
# if len(na_index) > 0:
# print('\nNon-structural positions detected for gene:', gene.lower()
# , '\nTotal number of these detected:', len(na_index)
# , '\These are at index:', na_index
# , '\nOriginal nrows:', len(my_df)
# , '\nDropping these...')
# my_df = my_df.drop(index=na_index)
# print('\nRevised nrows:', len(my_df))
# else:
# print('\nNo non-structural positions detected for gene:', gene.lower()
# , '\nnrows:', len(my_df))
###########################################################################
#%% Add lineage calculation columns
#FIXME: Check if this can be imported from config?
total_mtblineage_uc = 8
lineage_colnames = ['lineage_list_all', 'lineage_count_all', 'lineage_count_unique', 'lineage_list_unique', 'lineage_multimode']
#bar = my_df[lineage_colnames]
my_df['lineage_proportion'] = my_df['lineage_count_unique']/my_df['lineage_count_all']
my_df['dist_lineage_proportion'] = my_df['lineage_count_unique']/total_mtblineage_uc
###########################################################################
#%% Active site annotation column
# change from numberic to categorical
if my_df['active_site'].dtype in num_type:
my_df['active_site'] = my_df['active_site'].astype(object)
my_df['active_site'].dtype
#%% AA property change
#--------------------
# Water prop change
#--------------------
my_df['water_change'] = my_df['wt_prop_water'] + str('_to_') + my_df['mut_prop_water']
my_df['water_change'].value_counts()
water_prop_changeD = {
'hydrophobic_to_neutral' : 'change'
, 'hydrophobic_to_hydrophobic' : 'no_change'
, 'neutral_to_neutral' : 'no_change'
, 'neutral_to_hydrophobic' : 'change'
, 'hydrophobic_to_hydrophilic' : 'change'
, 'neutral_to_hydrophilic' : 'change'
, 'hydrophilic_to_neutral' : 'change'
, 'hydrophilic_to_hydrophobic' : 'change'
, 'hydrophilic_to_hydrophilic' : 'no_change'
}
my_df['water_change'] = my_df['water_change'].map(water_prop_changeD)
my_df['water_change'].value_counts()
#--------------------
# Polarity change
#--------------------
my_df['polarity_change'] = my_df['wt_prop_polarity'] + str('_to_') + my_df['mut_prop_polarity']
my_df['polarity_change'].value_counts()
polarity_prop_changeD = {
'non-polar_to_non-polar' : 'no_change'
, 'non-polar_to_neutral' : 'change'
, 'neutral_to_non-polar' : 'change'
, 'neutral_to_neutral' : 'no_change'
, 'non-polar_to_basic' : 'change'
, 'acidic_to_neutral' : 'change'
, 'basic_to_neutral' : 'change'
, 'non-polar_to_acidic' : 'change'
, 'neutral_to_basic' : 'change'
, 'acidic_to_non-polar' : 'change'
, 'basic_to_non-polar' : 'change'
, 'neutral_to_acidic' : 'change'
, 'acidic_to_acidic' : 'no_change'
, 'basic_to_acidic' : 'change'
, 'basic_to_basic' : 'no_change'
, 'acidic_to_basic' : 'change'}
my_df['polarity_change'] = my_df['polarity_change'].map(polarity_prop_changeD)
my_df['polarity_change'].value_counts()
#--------------------
# Electrostatics change
#--------------------
my_df['electrostatics_change'] = my_df['wt_calcprop'] + str('_to_') + my_df['mut_calcprop']
my_df['electrostatics_change'].value_counts()
calc_prop_changeD = {
'non-polar_to_non-polar' : 'no_change'
, 'non-polar_to_polar' : 'change'
, 'polar_to_non-polar' : 'change'
, 'non-polar_to_pos' : 'change'
, 'neg_to_non-polar' : 'change'
, 'non-polar_to_neg' : 'change'
, 'pos_to_polar' : 'change'
, 'pos_to_non-polar' : 'change'
, 'polar_to_polar' : 'no_change'
, 'neg_to_neg' : 'no_change'
, 'polar_to_neg' : 'change'
, 'pos_to_neg' : 'change'
, 'pos_to_pos' : 'no_change'
, 'polar_to_pos' : 'change'
, 'neg_to_polar' : 'change'
, 'neg_to_pos' : 'change'
}
my_df['electrostatics_change'] = my_df['electrostatics_change'].map(calc_prop_changeD)
my_df['electrostatics_change'].value_counts()
#--------------------
# Summary change: Create a combined column summarising these three cols
#--------------------
detect_change = 'change'
check_prop_cols = ['water_change', 'polarity_change', 'electrostatics_change']
#my_df['aa_prop_change'] = (my_df.values == detect_change).any(1).astype(int)
my_df['aa_prop_change'] = (my_df[check_prop_cols].values == detect_change).any(1).astype(int)
my_df['aa_prop_change'].value_counts()
my_df['aa_prop_change'].dtype
my_df['aa_prop_change'] = my_df['aa_prop_change'].map({1:'change'
, 0: 'no_change'})
my_df['aa_prop_change'].value_counts()
my_df['aa_prop_change'].dtype
#%% IMPUTE values for OR [check script for exploration: UQ_or_imputer]
#--------------------
# Impute OR values
#--------------------
#or_cols = ['or_mychisq', 'log10_or_mychisq', 'or_fisher']
sel_cols = ['mutationinformation', 'or_mychisq', 'log10_or_mychisq']
or_cols = ['or_mychisq', 'log10_or_mychisq']
print("count of NULL values before imputation\n")
print(my_df[or_cols].isnull().sum())
my_dfI = pd.DataFrame(index = my_df['mutationinformation'] )
my_dfI = pd.DataFrame(KNN(n_neighbors=3, weights="uniform").fit_transform(my_df[or_cols])
, index = my_df['mutationinformation']
, columns = or_cols )
my_dfI.columns = ['or_rawI', 'logorI']
my_dfI.columns
my_dfI = my_dfI.reset_index(drop = False) # prevents old index from being added as a column
my_dfI.head()
print("count of NULL values AFTER imputation\n")
print(my_dfI.isnull().sum())
#-------------------------------------------
# OR df Merge: with original based on index
#-------------------------------------------
#my_df['index_bm'] = my_df.index
mydf_imputed = pd.merge(my_df
, my_dfI
, on = 'mutationinformation')
#mydf_imputed = mydf_imputed.set_index(['index_bm'])
my_df['log10_or_mychisq'].isna().sum()
mydf_imputed['log10_or_mychisq'].isna().sum()
mydf_imputed['logorI'].isna().sum() # should be 0
len(my_df.columns)
len(mydf_imputed.columns)
#-----------------------------------------
# REASSIGN my_df after imputing OR values
#-----------------------------------------
my_df = mydf_imputed.copy()
if my_df['logorI'].isna().sum() == 0:
print('\nPASS: OR values imputed, data ready for ML')
else:
sys.exit('\nFAIL: something went wrong, Data not ready for ML. Please check upstream!')
#!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
#---------------------------------------
# TODO: try other imputation like MICE
#---------------------------------------
#!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
#%%########################################################################
#==========================
# Data for ML
#==========================
my_df_ml = my_df.copy()
# Build column names to mask for affinity chanhes
if gene.lower() in geneL_basic:
#X_stabilityN = common_cols_stabiltyN
gene_affinity_colnames = []# not needed as its the common ones
cols_to_mask = ['ligand_affinity_change']
if gene.lower() in geneL_ppi2:
gene_affinity_colnames = ['mcsm_ppi2_affinity', 'interface_dist']
#X_stabilityN = common_cols_stabiltyN + geneL_ppi2_st_cols
cols_to_mask = ['ligand_affinity_change', 'mcsm_ppi2_affinity']
if gene.lower() in geneL_na:
gene_affinity_colnames = ['mcsm_na_affinity']
#X_stabilityN = common_cols_stabiltyN + geneL_na_st_cols
cols_to_mask = ['ligand_affinity_change', 'mcsm_na_affinity']
if gene.lower() in geneL_na_ppi2:
gene_affinity_colnames = ['mcsm_na_affinity'] + ['mcsm_ppi2_affinity', 'interface_dist']
#X_stabilityN = common_cols_stabiltyN + geneL_na_ppi2_st_cols
cols_to_mask = ['ligand_affinity_change', 'mcsm_na_affinity', 'mcsm_ppi2_affinity']
#=======================
# Masking columns:
# (mCSM-lig, mCSM-NA, mCSM-ppi2) values for lig_dist >10
#=======================
my_df_ml['mutationinformation'][my_df_ml['ligand_distance']>10].value_counts()
my_df_ml.groupby('mutationinformation')['ligand_distance'].apply(lambda x: (x>10)).value_counts()
my_df_ml.loc[(my_df_ml['ligand_distance'] > 10), cols_to_mask].value_counts()
# mask the mcsm affinity related columns where ligand distance > 10
my_df_ml.loc[(my_df_ml['ligand_distance'] > 10), cols_to_mask] = 0
(my_df_ml['ligand_affinity_change'] == 0).sum()
mask_check = my_df_ml[['mutationinformation', 'ligand_distance'] + cols_to_mask]
#===================================================
# write file for check
mask_check.sort_values(by = ['ligand_distance'], ascending = True, inplace = True)
mask_check.to_csv(outdir + 'ml/' + gene.lower() + '_mask_check.csv')
#===================================================
###############################################################################
#%% Feature groups (FG): Build X for Input ML
############################################################################
#===========================
# FG1: Evolutionary features
#===========================
X_evolFN = ['consurf_score'
, 'snap2_score'
, 'provean_score']
###############################################################################
#========================
# FG2: Stability features
#========================
#--------
# common
#--------
X_common_stability_Fnum = [
'duet_stability_change'
, 'ddg_foldx'
, 'deepddg'
, 'ddg_dynamut2'
, 'contacts']
#--------
# FoldX
#--------
X_foldX_Fnum = [ 'electro_rr', 'electro_mm', 'electro_sm', 'electro_ss'
, 'disulfide_rr', 'disulfide_mm', 'disulfide_sm', 'disulfide_ss'
, 'hbonds_rr', 'hbonds_mm', 'hbonds_sm', 'hbonds_ss'
, 'partcov_rr', 'partcov_mm', 'partcov_sm', 'partcov_ss'
, 'vdwclashes_rr', 'vdwclashes_mm', 'vdwclashes_sm', 'vdwclashes_ss'
, 'volumetric_rr', 'volumetric_mm', 'volumetric_ss']
X_stability_FN = X_common_stability_Fnum + X_foldX_Fnum
###############################################################################
#===================
# FG3: Affinity features
#===================
common_affinity_Fnum = ['ligand_distance'
, 'ligand_affinity_change'
, 'mmcsm_lig']
# if gene.lower() in geneL_basic:
# X_affinityFN = common_affinity_Fnum
# else:
# X_affinityFN = common_affinity_Fnum + gene_affinity_colnames
X_affinityFN = common_affinity_Fnum + gene_affinity_colnames
###############################################################################
#============================
# FG4: Residue level features
#============================
#-----------
# AA index
#-----------
X_aaindex_Fnum = list(aa_df_cols)
print('\nTotal no. of features for aaindex:', len(X_aaindex_Fnum))
#-----------------
# surface area
# depth
# hydrophobicity
#-----------------
X_str_Fnum = ['rsa'
#, 'asa'
, 'kd_values'
, 'rd_values']
#---------------------------
# Other aa properties
# active site indication
#---------------------------
X_aap_Fcat = ['ss_class'
# , 'wt_prop_water'
# , 'mut_prop_water'
# , 'wt_prop_polarity'
# , 'mut_prop_polarity'
# , 'wt_calcprop'
# , 'mut_calcprop'
, 'aa_prop_change'
, 'electrostatics_change'
, 'polarity_change'
, 'water_change'
, 'active_site']
X_resprop_FN = X_aaindex_Fnum + X_str_Fnum + X_aap_Fcat
###############################################################################
#========================
# FG5: Genomic features
#========================
X_gn_mafor_Fnum = ['maf'
#, 'logorI'
# , 'or_rawI'
# , 'or_mychisq'
# , 'or_logistic'
# , 'or_fisher'
# , 'pval_fisher'
]
X_gn_linegae_Fnum = ['lineage_proportion'
, 'dist_lineage_proportion'
#, 'lineage' # could be included as a category but it has L2;L4 formatting
, 'lineage_count_all'
, 'lineage_count_unique'
]
# X_gn_Fcat = ['drtype_mode_labels' # beware then you can't use it to predict [USED it for uq_v1, not v2]
# #, 'gene_name' # will be required for the combined stuff
# ]
X_gn_Fcat = []
X_genomicFN = X_gn_mafor_Fnum + X_gn_linegae_Fnum + X_gn_Fcat
###############################################################################
#========================
# FG6 collapsed: Structural : Atability + Affinity + ResidueProp
#========================
X_structural_FN = X_stability_FN + X_affinityFN + X_resprop_FN
###############################################################################
#========================
# BUILDING all features
#========================
all_featuresN = X_evolFN + X_structural_FN + X_genomicFN
###############################################################################
#%% Define training and test data
#================================================================
# Training and BLIND test set: 80/20
# dst with actual values : training set
# dst with imputed values : THROW AWAY [unrepresentative]
#================================================================
my_df_ml[drug].isna().sum()
# blind_test_df = my_df_ml[my_df_ml[drug].isna()]
# blind_test_df.shape
training_df = my_df_ml[my_df_ml[drug].notna()]
training_df.shape
# Target 1: dst_mode
training_df[drug].value_counts()
training_df['dst_mode'].value_counts()
####################################################################
#====================================
# ML data: Train test split: 80/20
# with stratification
# 80% : training_data for CV
# 20% : blind test
#=====================================
x_features = training_df[all_featuresN]
y_target = training_df['dst_mode']
# sanity check
if not 'dst_mode' in x_features.columns:
print('\nPASS: x_features has no target variable')
x_ncols = len(x_features.columns)
print('\nNo. of columns for x_features:', x_ncols)
# NEED It for scaling law split
#https://towardsdatascience.com/finally-why-we-use-an-80-20-split-for-training-and-test-data-plus-an-alternative-method-oh-yes-edc77e96295d
else:
sys.exit('\nFAIL: x_features has target variable included. FIX it and rerun!')
#-------------------
# train-test split
#-------------------
#x_train, x_test, y_train, y_test # traditional var_names
# so my downstream code doesn't need to change
X, X_bts, y, y_bts = train_test_split(x_features, y_target
, test_size = 0.2
, **rs
, stratify = y_target)
yc1 = Counter(y)
yc1_ratio = yc1[0]/yc1[1]
yc2 = Counter(y_bts)
yc2_ratio = yc2[0]/yc2[1]
###############################################################################
#======================================================
# Determine categorical and numerical features
#======================================================
numerical_cols = X.select_dtypes(include=['int64', 'float64']).columns
numerical_cols
categorical_cols = X.select_dtypes(include=['object', 'bool']).columns
categorical_cols
################################################################################
# IMPORTANT sanity checks
if len(X.columns) == len(X_evolFN) + len(X_stability_FN) + len(X_affinityFN) + len(X_resprop_FN) + len(X_genomicFN):
print('\nPASS: ML data with input features, training and test generated...'
, '\n\nTotal no. of input features:' , len(X.columns)
, '\n--------No. of numerical features:' , len(numerical_cols)
, '\n--------No. of categorical features:' , len(categorical_cols)
, '\n\nTotal no. of evolutionary features:' , len(X_evolFN)
, '\n\nTotal no. of stability features:' , len(X_stability_FN)
, '\n--------Common stabilty cols:' , len(X_common_stability_Fnum)
, '\n--------Foldx cols:' , len(X_foldX_Fnum)
, '\n\nTotal no. of affinity features:' , len(X_affinityFN)
, '\n--------Common affinity cols:' , len(common_affinity_Fnum)
, '\n--------Gene specific affinity cols:' , len(gene_affinity_colnames)
, '\n\nTotal no. of residue level features:', len(X_resprop_FN)
, '\n--------AA index cols:' , len(X_aaindex_Fnum)
, '\n--------Residue Prop cols:' , len(X_str_Fnum)
, '\n--------AA change Prop cols:' , len(X_aap_Fcat)
, '\n\nTotal no. of genomic features:' , len(X_genomicFN)
, '\n--------MAF+OR cols:' , len(X_gn_mafor_Fnum)
, '\n--------Lineage cols:' , len(X_gn_linegae_Fnum)
, '\n--------Other cols:' , len(X_gn_Fcat)
)
else:
print('\nFAIL: numbers mismatch'
, '\nExpected:',len(X_evolFN) + len(X_stability_FN) + len(X_affinityFN) + len(X_resprop_FN) + len(X_genomicFN)
, '\nGot:', len(X.columns))
sys.exit()
###############################################################################
print('\n-------------------------------------------------------------'
, '\nSuccessfully split data: ALL features'
, '\nactual values: training set'
, '\nSplit:', tts_split
#, '\nimputed values: blind test set'
, '\n\nTotal data size:', len(X) + len(X_bts)
, '\n\nTrain data size:', X.shape
, '\ny_train numbers:', yc1
, '\n\nTest data size:', X_bts.shape
, '\ny_test_numbers:', yc2
, '\n\ny_train ratio:',yc1_ratio
, '\ny_test ratio:', yc2_ratio
, '\n-------------------------------------------------------------'
)
##########################################################################
# Quick check
#(X['ligand_affinity_change']==0).sum() == (X['ligand_distance']>10).sum()
for i in range(len(cols_to_mask)):
ind = i+1
print('\nindex:', i, '\nind:', ind)
print('\nMask count check:'
, (my_df_ml[cols_to_mask[i]]==0).sum() == (my_df_ml['ligand_distance']>10).sum()
)
print('Original Data\n', Counter(y)
, 'Data dim:', X.shape)
###########################################################################
#%%
###########################################################################
# RESAMPLING
###########################################################################
#------------------------------
# Simple Random oversampling
# [Numerical + catgeorical]
#------------------------------
oversample = RandomOverSampler(sampling_strategy='minority')
X_ros, y_ros = oversample.fit_resample(X, y)
print('\nSimple Random OverSampling\n', Counter(y_ros))
print(X_ros.shape)
#------------------------------
# Simple Random Undersampling
# [Numerical + catgeorical]
#------------------------------
undersample = RandomUnderSampler(sampling_strategy='majority')
X_rus, y_rus = undersample.fit_resample(X, y)
print('\nSimple Random UnderSampling\n', Counter(y_rus))
print(X_rus.shape)
#------------------------------
# Simple combine ROS and RUS
# [Numerical + catgeorical]
#------------------------------
oversample = RandomOverSampler(sampling_strategy='minority')
X_ros, y_ros = oversample.fit_resample(X, y)
undersample = RandomUnderSampler(sampling_strategy='majority')
X_rouC, y_rouC = undersample.fit_resample(X_ros, y_ros)
print('\nSimple Combined Over and UnderSampling\n', Counter(y_rouC))
print(X_rouC.shape)
#------------------------------
# SMOTE_NC: oversampling
# [numerical + categorical]
#https://stackoverflow.com/questions/47655813/oversampling-smote-for-binary-and-categorical-data-in-python
#------------------------------
# Determine categorical and numerical features
numerical_ix = X.select_dtypes(include=['int64', 'float64']).columns
numerical_ix
num_featuresL = list(numerical_ix)
numerical_colind = X.columns.get_indexer(list(numerical_ix) )
numerical_colind
categorical_ix = X.select_dtypes(include=['object', 'bool']).columns
categorical_ix
categorical_colind = X.columns.get_indexer(list(categorical_ix))
categorical_colind
k_sm = 5 # 5 is default
sm_nc = SMOTENC(categorical_features=categorical_colind, k_neighbors = k_sm, **rs, **njobs)
X_smnc, y_smnc = sm_nc.fit_resample(X, y)
print('\nSMOTE_NC OverSampling\n', Counter(y_smnc))
print(X_smnc.shape)
globals().update(locals()) # TROLOLOLOLOLOLS
#print("i did a horrible hack :-)")
###############################################################################
#%% SMOTE RESAMPLING for NUMERICAL ONLY*
# #------------------------------
# # SMOTE: Oversampling
# # [Numerical ONLY]
# #------------------------------
# k_sm = 1
# sm = SMOTE(sampling_strategy = 'auto', k_neighbors = k_sm, **rs)
# X_sm, y_sm = sm.fit_resample(X, y)
# print(X_sm.shape)
# print('\nSMOTE OverSampling\n', Counter(y_sm))
# y_sm_df = y_sm.to_frame()
# y_sm_df.value_counts().plot(kind = 'bar')
# #------------------------------
# # SMOTE: Over + Undersampling COMBINED
# # [Numerical ONLY]
# #-----------------------------
# sm_enn = SMOTEENN(enn=EditedNearestNeighbours(sampling_strategy='all', **rs, **njobs ))
# X_enn, y_enn = sm_enn.fit_resample(X, y)
# print(X_enn.shape)
# print('\nSMOTE Over+Under Sampling combined\n', Counter(y_enn))
###############################################################################
# TODO: Find over and undersampling JUST for categorical data
###########################################################################
print('\n#################################################################'
, '\nDim of X for gene:', gene.lower(), '\n', X.shape
, '\n###############################################################')

808
scripts/ml/ml_data_cd_7030.py
View file

@ -1,808 +0,0 @@
#!/usr/bin/env python3
# -*- coding: utf-8 -*-
"""
Created on Sun Mar 6 13:41:54 2022
@author: tanu
"""
def setvars(gene,drug):
#https://stackoverflow.com/questions/51695322/compare-multiple-algorithms-with-sklearn-pipeline
import os, sys
import pandas as pd
import numpy as np
print(np.__version__)
print(pd.__version__)
import pprint as pp
from copy import deepcopy
from collections import Counter
from sklearn.impute import KNNImputer as KNN
from imblearn.over_sampling import RandomOverSampler
from imblearn.under_sampling import RandomUnderSampler
from imblearn.over_sampling import SMOTE
from sklearn.datasets import make_classification
from imblearn.combine import SMOTEENN
from imblearn.combine import SMOTETomek
from imblearn.over_sampling import SMOTENC
from imblearn.under_sampling import EditedNearestNeighbours
from imblearn.under_sampling import RepeatedEditedNearestNeighbours
from sklearn.metrics import make_scorer, confusion_matrix, accuracy_score, balanced_accuracy_score, precision_score, average_precision_score, recall_score
from sklearn.metrics import roc_auc_score, roc_curve, f1_score, matthews_corrcoef, jaccard_score, classification_report
from sklearn.model_selection import train_test_split, cross_validate, cross_val_score
from sklearn.model_selection import StratifiedKFold,RepeatedStratifiedKFold, RepeatedKFold
from sklearn.pipeline import Pipeline, make_pipeline
import argparse
import re
#%% GLOBALS
tts_split = "70_30"
rs = {'random_state': 42}
njobs = {'n_jobs': 10}
scoring_fn = ({ 'mcc' : make_scorer(matthews_corrcoef)
, 'accuracy' : make_scorer(accuracy_score)
, 'fscore' : make_scorer(f1_score)
, 'precision' : make_scorer(precision_score)
, 'recall' : make_scorer(recall_score)
, 'roc_auc' : make_scorer(roc_auc_score)
, 'jcc' : make_scorer(jaccard_score)
})
skf_cv = StratifiedKFold(n_splits = 10
#, shuffle = False, random_state= None)
, shuffle = True,**rs)
rskf_cv = RepeatedStratifiedKFold(n_splits = 10
, n_repeats = 3
, **rs)
mcc_score_fn = {'mcc': make_scorer(matthews_corrcoef)}
jacc_score_fn = {'jcc': make_scorer(jaccard_score)}
#%% FOR LATER: Combine ED logo data
###########################################################################
homedir = os.path.expanduser("~")
geneL_basic = ['pnca']
geneL_na = ['gid']
geneL_na_ppi2 = ['rpob']
geneL_ppi2 = ['alr', 'embb', 'katg']
#num_type = ['int64', 'float64']
num_type = ['int16', 'int32', 'int64', 'float16', 'float32', 'float64']
cat_type = ['object', 'bool']
#==============
# directories
#==============
datadir = homedir + '/git/Data/'
indir = datadir + drug + '/input/'
outdir = datadir + drug + '/output/'
#=======
# input
#=======
#---------
# File 1
#---------
infile_ml1 = outdir + gene.lower() + '_merged_df3.csv'
#infile_ml2 = outdir + gene.lower() + '_merged_df2.csv'
my_features_df = pd.read_csv(infile_ml1, index_col = 0)
my_features_df = my_features_df .reset_index(drop = True)
my_features_df.index
my_features_df.dtypes
mycols = my_features_df.columns
#---------
# File 2
#---------
infile_aaindex = outdir + 'aa_index/' + gene.lower() + '_aa.csv'
aaindex_df = pd.read_csv(infile_aaindex, index_col = 0)
aaindex_df.dtypes
#-----------
# check for non-numerical columns
#-----------
if any(aaindex_df.dtypes==object):
print('\naaindex_df contains non-numerical data')
aaindex_df_object = aaindex_df.select_dtypes(include = cat_type)
print('\nTotal no. of non-numerial columns:', len(aaindex_df_object.columns))
expected_aa_ncols = len(aaindex_df.columns) - len(aaindex_df_object.columns)
#-----------
# Extract numerical data only
#-----------
print('\nSelecting numerical data only')
aaindex_df = aaindex_df.select_dtypes(include = num_type)
#---------------------------
# aaindex: sanity check 1
#---------------------------
if len(aaindex_df.columns) == expected_aa_ncols:
print('\nPASS: successfully selected numerical columns only for aaindex_df')
else:
print('\nFAIL: Numbers mismatch'
, '\nExpected ncols:', expected_aa_ncols
, '\nGot:', len(aaindex_df.columns))
#---------------
# check for NA
#---------------
print('\nNow checking for NA in the remaining aaindex_cols')
c1 = aaindex_df.isna().sum()
c2 = c1.sort_values(ascending=False)
print('\nCounting aaindex_df cols with NA'
, '\nncols with NA:', sum(c2>0), 'columns'
, '\nDropping these...'
, '\nOriginal ncols:', len(aaindex_df.columns)
)
aa_df = aaindex_df.dropna(axis=1)
print('\nRevised df ncols:', len(aa_df.columns))
c3 = aa_df.isna().sum()
c4 = c3.sort_values(ascending=False)
print('\nChecking NA in revised df...')
if sum(c4>0):
sys.exit('\nFAIL: aaindex_df still contains cols with NA, please check and drop these before proceeding...')
else:
print('\nPASS: cols with NA successfully dropped from aaindex_df'
, '\nProceeding with combining aa_df with other features_df')
#---------------------------
# aaindex: sanity check 2
#---------------------------
expected_aa_ncols2 = len(aaindex_df.columns) - sum(c2>0)
if len(aa_df.columns) == expected_aa_ncols2:
print('\nPASS: ncols match'
, '\nExpected ncols:', expected_aa_ncols2
, '\nGot:', len(aa_df.columns))
else:
print('\nFAIL: Numbers mismatch'
, '\nExpected ncols:', expected_aa_ncols2
, '\nGot:', len(aa_df.columns))
# Important: need this to identify aaindex cols
aa_df_cols = aa_df.columns
print('\nTotal no. of columns in clean aa_df:', len(aa_df_cols))
###############################################################################
#%% Combining my_features_df and aaindex_df
#===========================
# Merge my_df + aaindex_df
#===========================
if aa_df.columns[aa_df.columns.isin(my_features_df.columns)] == my_features_df.columns[my_features_df.columns.isin(aa_df.columns)]:
print('\nMerging on column: mutationinformation')
if len(my_features_df) == len(aa_df):
expected_nrows = len(my_features_df)
print('\nProceeding to merge, expected nrows in merged_df:', expected_nrows)
else:
sys.exit('\nNrows mismatch, cannot merge. Please check'
, '\nnrows my_df:', len(my_features_df)
, '\nnrows aa_df:', len(aa_df))
#-----------------
# Reset index: mutationinformation
# Very important for merging
#-----------------
aa_df = aa_df.reset_index()
expected_ncols = len(my_features_df.columns) + len(aa_df.columns) - 1 # for the no. of merging col
#-----------------
# Merge: my_features_df + aa_df
#-----------------
merged_df = pd.merge(my_features_df
, aa_df
, on = 'mutationinformation')
#---------------------------
# aaindex: sanity check 3
#---------------------------
if len(merged_df.columns) == expected_ncols:
print('\nPASS: my_features_df and aa_df successfully combined'
, '\nnrows:', len(merged_df)
, '\nncols:', len(merged_df.columns))
else:
sys.exit('\nFAIL: could not combine my_features_df and aa_df'
, '\nCheck dims and merging cols!')
#--------
# Reassign so downstream code doesn't need to change
#--------
my_df = merged_df.copy()
#%% Data: my_df
# Check if non structural pos have crept in
# IDEALLY remove from source! But for rpoB do it here
# Drop NA where numerical cols have them
if gene.lower() in geneL_na_ppi2:
#D1148 get rid of
na_index = my_df['mutationinformation'].index[my_df['mcsm_na_affinity'].apply(np.isnan)]
my_df = my_df.drop(index=na_index)
# FIXED: complete data for all muts inc L114M, F115L, V123L, V125I, V131M
# if gene.lower() in ['embb']:
# na_index = my_df['mutationinformation'].index[my_df['ligand_distance'].apply(np.isnan)]
# my_df = my_df.drop(index=na_index)
# # Sanity check for non-structural positions
# print('\nChecking for non-structural postions')
# na_index = my_df['mutationinformation'].index[my_df['ligand_distance'].apply(np.isnan)]
# if len(na_index) > 0:
# print('\nNon-structural positions detected for gene:', gene.lower()
# , '\nTotal number of these detected:', len(na_index)
# , '\These are at index:', na_index
# , '\nOriginal nrows:', len(my_df)
# , '\nDropping these...')
# my_df = my_df.drop(index=na_index)
# print('\nRevised nrows:', len(my_df))
# else:
# print('\nNo non-structural positions detected for gene:', gene.lower()
# , '\nnrows:', len(my_df))
###########################################################################
#%% Add lineage calculation columns
#FIXME: Check if this can be imported from config?
total_mtblineage_uc = 8
lineage_colnames = ['lineage_list_all', 'lineage_count_all', 'lineage_count_unique', 'lineage_list_unique', 'lineage_multimode']
#bar = my_df[lineage_colnames]
my_df['lineage_proportion'] = my_df['lineage_count_unique']/my_df['lineage_count_all']
my_df['dist_lineage_proportion'] = my_df['lineage_count_unique']/total_mtblineage_uc
###########################################################################
#%% Active site annotation column
# change from numberic to categorical
if my_df['active_site'].dtype in num_type:
my_df['active_site'] = my_df['active_site'].astype(object)
my_df['active_site'].dtype
#%% AA property change
#--------------------
# Water prop change
#--------------------
my_df['water_change'] = my_df['wt_prop_water'] + str('_to_') + my_df['mut_prop_water']
my_df['water_change'].value_counts()
water_prop_changeD = {
'hydrophobic_to_neutral' : 'change'
, 'hydrophobic_to_hydrophobic' : 'no_change'
, 'neutral_to_neutral' : 'no_change'
, 'neutral_to_hydrophobic' : 'change'
, 'hydrophobic_to_hydrophilic' : 'change'
, 'neutral_to_hydrophilic' : 'change'
, 'hydrophilic_to_neutral' : 'change'
, 'hydrophilic_to_hydrophobic' : 'change'
, 'hydrophilic_to_hydrophilic' : 'no_change'
}
my_df['water_change'] = my_df['water_change'].map(water_prop_changeD)
my_df['water_change'].value_counts()
#--------------------
# Polarity change
#--------------------
my_df['polarity_change'] = my_df['wt_prop_polarity'] + str('_to_') + my_df['mut_prop_polarity']
my_df['polarity_change'].value_counts()
polarity_prop_changeD = {
'non-polar_to_non-polar' : 'no_change'
, 'non-polar_to_neutral' : 'change'
, 'neutral_to_non-polar' : 'change'
, 'neutral_to_neutral' : 'no_change'
, 'non-polar_to_basic' : 'change'
, 'acidic_to_neutral' : 'change'
, 'basic_to_neutral' : 'change'
, 'non-polar_to_acidic' : 'change'
, 'neutral_to_basic' : 'change'
, 'acidic_to_non-polar' : 'change'
, 'basic_to_non-polar' : 'change'
, 'neutral_to_acidic' : 'change'
, 'acidic_to_acidic' : 'no_change'
, 'basic_to_acidic' : 'change'
, 'basic_to_basic' : 'no_change'
, 'acidic_to_basic' : 'change'}
my_df['polarity_change'] = my_df['polarity_change'].map(polarity_prop_changeD)
my_df['polarity_change'].value_counts()
#--------------------
# Electrostatics change
#--------------------
my_df['electrostatics_change'] = my_df['wt_calcprop'] + str('_to_') + my_df['mut_calcprop']
my_df['electrostatics_change'].value_counts()
calc_prop_changeD = {
'non-polar_to_non-polar' : 'no_change'
, 'non-polar_to_polar' : 'change'
, 'polar_to_non-polar' : 'change'
, 'non-polar_to_pos' : 'change'
, 'neg_to_non-polar' : 'change'
, 'non-polar_to_neg' : 'change'
, 'pos_to_polar' : 'change'
, 'pos_to_non-polar' : 'change'
, 'polar_to_polar' : 'no_change'
, 'neg_to_neg' : 'no_change'
, 'polar_to_neg' : 'change'
, 'pos_to_neg' : 'change'
, 'pos_to_pos' : 'no_change'
, 'polar_to_pos' : 'change'
, 'neg_to_polar' : 'change'
, 'neg_to_pos' : 'change'
}
my_df['electrostatics_change'] = my_df['electrostatics_change'].map(calc_prop_changeD)
my_df['electrostatics_change'].value_counts()
#--------------------
# Summary change: Create a combined column summarising these three cols
#--------------------
detect_change = 'change'
check_prop_cols = ['water_change', 'polarity_change', 'electrostatics_change']
#my_df['aa_prop_change'] = (my_df.values == detect_change).any(1).astype(int)
my_df['aa_prop_change'] = (my_df[check_prop_cols].values == detect_change).any(1).astype(int)
my_df['aa_prop_change'].value_counts()
my_df['aa_prop_change'].dtype
my_df['aa_prop_change'] = my_df['aa_prop_change'].map({1:'change'
, 0: 'no_change'})
my_df['aa_prop_change'].value_counts()
my_df['aa_prop_change'].dtype
#%% IMPUTE values for OR [check script for exploration: UQ_or_imputer]
#--------------------
# Impute OR values
#--------------------
#or_cols = ['or_mychisq', 'log10_or_mychisq', 'or_fisher']
sel_cols = ['mutationinformation', 'or_mychisq', 'log10_or_mychisq']
or_cols = ['or_mychisq', 'log10_or_mychisq']
print("count of NULL values before imputation\n")
print(my_df[or_cols].isnull().sum())
my_dfI = pd.DataFrame(index = my_df['mutationinformation'] )
my_dfI = pd.DataFrame(KNN(n_neighbors=3, weights="uniform").fit_transform(my_df[or_cols])
, index = my_df['mutationinformation']
, columns = or_cols )
my_dfI.columns = ['or_rawI', 'logorI']
my_dfI.columns
my_dfI = my_dfI.reset_index(drop = False) # prevents old index from being added as a column
my_dfI.head()
print("count of NULL values AFTER imputation\n")
print(my_dfI.isnull().sum())
#-------------------------------------------
# OR df Merge: with original based on index
#-------------------------------------------
#my_df['index_bm'] = my_df.index
mydf_imputed = pd.merge(my_df
, my_dfI
, on = 'mutationinformation')
#mydf_imputed = mydf_imputed.set_index(['index_bm'])
my_df['log10_or_mychisq'].isna().sum()
mydf_imputed['log10_or_mychisq'].isna().sum()
mydf_imputed['logorI'].isna().sum() # should be 0
len(my_df.columns)
len(mydf_imputed.columns)
#-----------------------------------------
# REASSIGN my_df after imputing OR values
#-----------------------------------------
my_df = mydf_imputed.copy()
if my_df['logorI'].isna().sum() == 0:
print('\nPASS: OR values imputed, data ready for ML')
else:
sys.exit('\nFAIL: something went wrong, Data not ready for ML. Please check upstream!')
#!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
#---------------------------------------
# TODO: try other imputation like MICE
#---------------------------------------
#!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
#%%########################################################################
#==========================
# Data for ML
#==========================
my_df_ml = my_df.copy()
# Build column names to mask for affinity chanhes
if gene.lower() in geneL_basic:
#X_stabilityN = common_cols_stabiltyN
gene_affinity_colnames = []# not needed as its the common ones
cols_to_mask = ['ligand_affinity_change']
if gene.lower() in geneL_ppi2:
gene_affinity_colnames = ['mcsm_ppi2_affinity', 'interface_dist']
#X_stabilityN = common_cols_stabiltyN + geneL_ppi2_st_cols
cols_to_mask = ['ligand_affinity_change', 'mcsm_ppi2_affinity']
if gene.lower() in geneL_na:
gene_affinity_colnames = ['mcsm_na_affinity']
#X_stabilityN = common_cols_stabiltyN + geneL_na_st_cols
cols_to_mask = ['ligand_affinity_change', 'mcsm_na_affinity']
if gene.lower() in geneL_na_ppi2:
gene_affinity_colnames = ['mcsm_na_affinity'] + ['mcsm_ppi2_affinity', 'interface_dist']
#X_stabilityN = common_cols_stabiltyN + geneL_na_ppi2_st_cols
cols_to_mask = ['ligand_affinity_change', 'mcsm_na_affinity', 'mcsm_ppi2_affinity']
#=======================
# Masking columns:
# (mCSM-lig, mCSM-NA, mCSM-ppi2) values for lig_dist >10
#=======================
my_df_ml['mutationinformation'][my_df_ml['ligand_distance']>10].value_counts()
my_df_ml.groupby('mutationinformation')['ligand_distance'].apply(lambda x: (x>10)).value_counts()
my_df_ml.loc[(my_df_ml['ligand_distance'] > 10), cols_to_mask].value_counts()
# mask the mcsm affinity related columns where ligand distance > 10
my_df_ml.loc[(my_df_ml['ligand_distance'] > 10), cols_to_mask] = 0
(my_df_ml['ligand_affinity_change'] == 0).sum()
mask_check = my_df_ml[['mutationinformation', 'ligand_distance'] + cols_to_mask]
#===================================================
# write file for check
mask_check.sort_values(by = ['ligand_distance'], ascending = True, inplace = True)
mask_check.to_csv(outdir + 'ml/' + gene.lower() + '_mask_check.csv')
#===================================================
###############################################################################
#%% Feature groups (FG): Build X for Input ML
############################################################################
#===========================
# FG1: Evolutionary features
#===========================
X_evolFN = ['consurf_score'
, 'snap2_score'
, 'provean_score']
###############################################################################
#========================
# FG2: Stability features
#========================
#--------
# common
#--------
X_common_stability_Fnum = [
'duet_stability_change'
, 'ddg_foldx'
, 'deepddg'
, 'ddg_dynamut2'
, 'contacts']
#--------
# FoldX
#--------
X_foldX_Fnum = [ 'electro_rr', 'electro_mm', 'electro_sm', 'electro_ss'
, 'disulfide_rr', 'disulfide_mm', 'disulfide_sm', 'disulfide_ss'
, 'hbonds_rr', 'hbonds_mm', 'hbonds_sm', 'hbonds_ss'
, 'partcov_rr', 'partcov_mm', 'partcov_sm', 'partcov_ss'
, 'vdwclashes_rr', 'vdwclashes_mm', 'vdwclashes_sm', 'vdwclashes_ss'
, 'volumetric_rr', 'volumetric_mm', 'volumetric_ss']
X_stability_FN = X_common_stability_Fnum + X_foldX_Fnum
###############################################################################
#===================
# FG3: Affinity features
#===================
common_affinity_Fnum = ['ligand_distance'
, 'ligand_affinity_change'
, 'mmcsm_lig']
# if gene.lower() in geneL_basic:
# X_affinityFN = common_affinity_Fnum
# else:
# X_affinityFN = common_affinity_Fnum + gene_affinity_colnames
X_affinityFN = common_affinity_Fnum + gene_affinity_colnames
###############################################################################
#============================
# FG4: Residue level features
#============================
#-----------
# AA index
#-----------
X_aaindex_Fnum = list(aa_df_cols)
print('\nTotal no. of features for aaindex:', len(X_aaindex_Fnum))
#-----------------
# surface area
# depth
# hydrophobicity
#-----------------
X_str_Fnum = ['rsa'
#, 'asa'
, 'kd_values'
, 'rd_values']
#---------------------------
# Other aa properties
# active site indication
#---------------------------
X_aap_Fcat = ['ss_class'
# , 'wt_prop_water'
# , 'mut_prop_water'
# , 'wt_prop_polarity'
# , 'mut_prop_polarity'
# , 'wt_calcprop'
# , 'mut_calcprop'
, 'aa_prop_change'
, 'electrostatics_change'
, 'polarity_change'
, 'water_change'
, 'active_site']
X_resprop_FN = X_aaindex_Fnum + X_str_Fnum + X_aap_Fcat
###############################################################################
#========================
# FG5: Genomic features
#========================
X_gn_mafor_Fnum = ['maf'
#, 'logorI'
# , 'or_rawI'
# , 'or_mychisq'
# , 'or_logistic'
# , 'or_fisher'
# , 'pval_fisher'
]
X_gn_linegae_Fnum = ['lineage_proportion'
, 'dist_lineage_proportion'
#, 'lineage' # could be included as a category but it has L2;L4 formatting
, 'lineage_count_all'
, 'lineage_count_unique'
]
# X_gn_Fcat = ['drtype_mode_labels' # beware then you can't use it to predict [USED it for uq_v1, not v2]
# #, 'gene_name' # will be required for the combined stuff
# ]
X_gn_Fcat = []
X_genomicFN = X_gn_mafor_Fnum + X_gn_linegae_Fnum + X_gn_Fcat
###############################################################################
#========================
# FG6 collapsed: Structural : Atability + Affinity + ResidueProp
#========================
X_structural_FN = X_stability_FN + X_affinityFN + X_resprop_FN
###############################################################################
#========================
# BUILDING all features
#========================
all_featuresN = X_evolFN + X_structural_FN + X_genomicFN
###############################################################################
#%% Define training and test data
#================================================================
# Training and BLIND test set: 70/30
# dst with actual values : training set
# dst with imputed values : THROW AWAY [unrepresentative]
#================================================================
my_df_ml[drug].isna().sum()
# blind_test_df = my_df_ml[my_df_ml[drug].isna()]
# blind_test_df.shape
#training_df = my_df_ml[my_df_ml[drug].notna()]
#training_df.shape
training_df = my_df_ml.copy()
# Target 1: dst_mode
training_df[drug].value_counts()
training_df['dst_mode'].value_counts()
####################################################################
#====================================
# ML data: Train test split: 70/30
# with stratification
# 70% : training_data for CV
# 30% : blind test
#=====================================
x_features = training_df[all_featuresN]
y_target = training_df['dst_mode']
# sanity check
if not 'dst_mode' in x_features.columns:
print('\nPASS: x_features has no target variable')
x_ncols = len(x_features.columns)
print('\nNo. of columns for x_features:', x_ncols)
# NEED It for scaling law split
#https://towardsdatascience.com/finally-why-we-use-an-80-20-split-for-training-and-test-data-plus-an-alternative-method-oh-yes-edc77e96295d
else:
sys.exit('\nFAIL: x_features has target variable included. FIX it and rerun!')
#-------------------
# train-test split
#-------------------
#x_train, x_test, y_train, y_test # traditional var_names
# so my downstream code doesn't need to change
X, X_bts, y, y_bts = train_test_split(x_features, y_target
, test_size = 0.33
, **rs
, stratify = y_target)
yc1 = Counter(y)
yc1_ratio = yc1[0]/yc1[1]
yc2 = Counter(y_bts)
yc2_ratio = yc2[0]/yc2[1]
###############################################################################
#======================================================
# Determine categorical and numerical features
#======================================================
numerical_cols = X.select_dtypes(include=['int64', 'float64']).columns
numerical_cols
categorical_cols = X.select_dtypes(include=['object', 'bool']).columns
categorical_cols
################################################################################
# IMPORTANT sanity checks
if len(X.columns) == len(X_evolFN) + len(X_stability_FN) + len(X_affinityFN) + len(X_resprop_FN) + len(X_genomicFN):
print('\nPASS: ML data with input features, training and test generated...'
, '\n\nTotal no. of input features:' , len(X.columns)
, '\n--------No. of numerical features:' , len(numerical_cols)
, '\n--------No. of categorical features:' , len(categorical_cols)
, '\n\nTotal no. of evolutionary features:' , len(X_evolFN)
, '\n\nTotal no. of stability features:' , len(X_stability_FN)
, '\n--------Common stabilty cols:' , len(X_common_stability_Fnum)
, '\n--------Foldx cols:' , len(X_foldX_Fnum)
, '\n\nTotal no. of affinity features:' , len(X_affinityFN)
, '\n--------Common affinity cols:' , len(common_affinity_Fnum)
, '\n--------Gene specific affinity cols:' , len(gene_affinity_colnames)
, '\n\nTotal no. of residue level features:', len(X_resprop_FN)
, '\n--------AA index cols:' , len(X_aaindex_Fnum)
, '\n--------Residue Prop cols:' , len(X_str_Fnum)
, '\n--------AA change Prop cols:' , len(X_aap_Fcat)
, '\n\nTotal no. of genomic features:' , len(X_genomicFN)
, '\n--------MAF+OR cols:' , len(X_gn_mafor_Fnum)
, '\n--------Lineage cols:' , len(X_gn_linegae_Fnum)
, '\n--------Other cols:' , len(X_gn_Fcat)
)
else:
print('\nFAIL: numbers mismatch'
, '\nExpected:',len(X_evolFN) + len(X_stability_FN) + len(X_affinityFN) + len(X_resprop_FN) + len(X_genomicFN)
, '\nGot:', len(X.columns))
sys.exit()
###############################################################################
print('\n-------------------------------------------------------------'
, '\nSuccessfully split data: ALL features'
, '\nactual values: training set'
, '\nSplit:', tts_split
#, '\nimputed values: blind test set'
, '\n\nTotal data size:', len(X) + len(X_bts)
, '\n\nTrain data size:', X.shape
, '\ny_train numbers:', yc1
, '\n\nTest data size:', X_bts.shape
, '\ny_test_numbers:', yc2
, '\n\ny_train ratio:',yc1_ratio
, '\ny_test ratio:', yc2_ratio
, '\n-------------------------------------------------------------'
)
##########################################################################
# Quick check
#(X['ligand_affinity_change']==0).sum() == (X['ligand_distance']>10).sum()
for i in range(len(cols_to_mask)):
ind = i+1
print('\nindex:', i, '\nind:', ind)
print('\nMask count check:'
, (my_df_ml[cols_to_mask[i]]==0).sum() == (my_df_ml['ligand_distance']>10).sum()
)
print('Original Data\n', Counter(y)
, 'Data dim:', X.shape)
###########################################################################
#%%
###########################################################################
# RESAMPLING
###########################################################################
#------------------------------
# Simple Random oversampling
# [Numerical + catgeorical]
#------------------------------
oversample = RandomOverSampler(sampling_strategy='minority')
X_ros, y_ros = oversample.fit_resample(X, y)
print('\nSimple Random OverSampling\n', Counter(y_ros))
print(X_ros.shape)
#------------------------------
# Simple Random Undersampling
# [Numerical + catgeorical]
#------------------------------
undersample = RandomUnderSampler(sampling_strategy='majority')
X_rus, y_rus = undersample.fit_resample(X, y)
print('\nSimple Random UnderSampling\n', Counter(y_rus))
print(X_rus.shape)
#------------------------------
# Simple combine ROS and RUS
# [Numerical + catgeorical]
#------------------------------
oversample = RandomOverSampler(sampling_strategy='minority')
X_ros, y_ros = oversample.fit_resample(X, y)
undersample = RandomUnderSampler(sampling_strategy='majority')
X_rouC, y_rouC = undersample.fit_resample(X_ros, y_ros)
print('\nSimple Combined Over and UnderSampling\n', Counter(y_rouC))
print(X_rouC.shape)
#------------------------------
# SMOTE_NC: oversampling
# [numerical + categorical]
#https://stackoverflow.com/questions/47655813/oversampling-smote-for-binary-and-categorical-data-in-python
#------------------------------
# Determine categorical and numerical features
numerical_ix = X.select_dtypes(include=['int64', 'float64']).columns
numerical_ix
num_featuresL = list(numerical_ix)
numerical_colind = X.columns.get_indexer(list(numerical_ix) )
numerical_colind
categorical_ix = X.select_dtypes(include=['object', 'bool']).columns
categorical_ix
categorical_colind = X.columns.get_indexer(list(categorical_ix))
categorical_colind
k_sm = 5 # 5 is default
sm_nc = SMOTENC(categorical_features=categorical_colind, k_neighbors = k_sm, **rs, **njobs)
X_smnc, y_smnc = sm_nc.fit_resample(X, y)
print('\nSMOTE_NC OverSampling\n', Counter(y_smnc))
print(X_smnc.shape)
globals().update(locals()) # TROLOLOLOLOLOLS
#print("i did a horrible hack :-)")
###############################################################################
#%% SMOTE RESAMPLING for NUMERICAL ONLY*
# #------------------------------
# # SMOTE: Oversampling
# # [Numerical ONLY]
# #------------------------------
# k_sm = 1
# sm = SMOTE(sampling_strategy = 'auto', k_neighbors = k_sm, **rs)
# X_sm, y_sm = sm.fit_resample(X, y)
# print(X_sm.shape)
# print('\nSMOTE OverSampling\n', Counter(y_sm))
# y_sm_df = y_sm.to_frame()
# y_sm_df.value_counts().plot(kind = 'bar')
# #------------------------------
# # SMOTE: Over + Undersampling COMBINED
# # [Numerical ONLY]
# #-----------------------------
# sm_enn = SMOTEENN(enn=EditedNearestNeighbours(sampling_strategy='all', **rs, **njobs ))
# X_enn, y_enn = sm_enn.fit_resample(X, y)
# print(X_enn.shape)
# print('\nSMOTE Over+Under Sampling combined\n', Counter(y_enn))
###############################################################################
# TODO: Find over and undersampling JUST for categorical data
###########################################################################
print('\n#################################################################'
, '\nDim of X for gene:', gene.lower(), '\n', X.shape
, '\n###############################################################')

808
scripts/ml/ml_data_cd_8020.py
View file

@ -1,808 +0,0 @@
#!/usr/bin/env python3
# -*- coding: utf-8 -*-
"""
Created on Sun Mar 6 13:41:54 2022
@author: tanu
"""
def setvars(gene,drug):
#https://stackoverflow.com/questions/51695322/compare-multiple-algorithms-with-sklearn-pipeline
import os, sys
import pandas as pd
import numpy as np
print(np.__version__)
print(pd.__version__)
import pprint as pp
from copy import deepcopy
from collections import Counter
from sklearn.impute import KNNImputer as KNN
from imblearn.over_sampling import RandomOverSampler
from imblearn.under_sampling import RandomUnderSampler
from imblearn.over_sampling import SMOTE
from sklearn.datasets import make_classification
from imblearn.combine import SMOTEENN
from imblearn.combine import SMOTETomek
from imblearn.over_sampling import SMOTENC
from imblearn.under_sampling import EditedNearestNeighbours
from imblearn.under_sampling import RepeatedEditedNearestNeighbours
from sklearn.metrics import make_scorer, confusion_matrix, accuracy_score, balanced_accuracy_score, precision_score, average_precision_score, recall_score
from sklearn.metrics import roc_auc_score, roc_curve, f1_score, matthews_corrcoef, jaccard_score, classification_report
from sklearn.model_selection import train_test_split, cross_validate, cross_val_score
from sklearn.model_selection import StratifiedKFold,RepeatedStratifiedKFold, RepeatedKFold
from sklearn.pipeline import Pipeline, make_pipeline
import argparse
import re
#%% GLOBALS
tts_split = "80_20"
rs = {'random_state': 42}
njobs = {'n_jobs': 10}
scoring_fn = ({ 'mcc' : make_scorer(matthews_corrcoef)
, 'accuracy' : make_scorer(accuracy_score)
, 'fscore' : make_scorer(f1_score)
, 'precision' : make_scorer(precision_score)
, 'recall' : make_scorer(recall_score)
, 'roc_auc' : make_scorer(roc_auc_score)
, 'jcc' : make_scorer(jaccard_score)
})
skf_cv = StratifiedKFold(n_splits = 10
#, shuffle = False, random_state= None)
, shuffle = True,**rs)
rskf_cv = RepeatedStratifiedKFold(n_splits = 10
, n_repeats = 3
, **rs)
mcc_score_fn = {'mcc': make_scorer(matthews_corrcoef)}
jacc_score_fn = {'jcc': make_scorer(jaccard_score)}
#%% FOR LATER: Combine ED logo data
###########################################################################
homedir = os.path.expanduser("~")
geneL_basic = ['pnca']
geneL_na = ['gid']
geneL_na_ppi2 = ['rpob']
geneL_ppi2 = ['alr', 'embb', 'katg']
#num_type = ['int64', 'float64']
num_type = ['int16', 'int32', 'int64', 'float16', 'float32', 'float64']
cat_type = ['object', 'bool']
#==============
# directories
#==============
datadir = homedir + '/git/Data/'
indir = datadir + drug + '/input/'
outdir = datadir + drug + '/output/'
#=======
# input
#=======
#---------
# File 1
#---------
infile_ml1 = outdir + gene.lower() + '_merged_df3.csv'
#infile_ml2 = outdir + gene.lower() + '_merged_df2.csv'
my_features_df = pd.read_csv(infile_ml1, index_col = 0)
my_features_df = my_features_df .reset_index(drop = True)
my_features_df.index
my_features_df.dtypes
mycols = my_features_df.columns
#---------
# File 2
#---------
infile_aaindex = outdir + 'aa_index/' + gene.lower() + '_aa.csv'
aaindex_df = pd.read_csv(infile_aaindex, index_col = 0)
aaindex_df.dtypes
#-----------
# check for non-numerical columns
#-----------
if any(aaindex_df.dtypes==object):
print('\naaindex_df contains non-numerical data')
aaindex_df_object = aaindex_df.select_dtypes(include = cat_type)
print('\nTotal no. of non-numerial columns:', len(aaindex_df_object.columns))
expected_aa_ncols = len(aaindex_df.columns) - len(aaindex_df_object.columns)
#-----------
# Extract numerical data only
#-----------
print('\nSelecting numerical data only')
aaindex_df = aaindex_df.select_dtypes(include = num_type)
#---------------------------
# aaindex: sanity check 1
#---------------------------
if len(aaindex_df.columns) == expected_aa_ncols:
print('\nPASS: successfully selected numerical columns only for aaindex_df')
else:
print('\nFAIL: Numbers mismatch'
, '\nExpected ncols:', expected_aa_ncols
, '\nGot:', len(aaindex_df.columns))
#---------------
# check for NA
#---------------
print('\nNow checking for NA in the remaining aaindex_cols')
c1 = aaindex_df.isna().sum()
c2 = c1.sort_values(ascending=False)
print('\nCounting aaindex_df cols with NA'
, '\nncols with NA:', sum(c2>0), 'columns'
, '\nDropping these...'
, '\nOriginal ncols:', len(aaindex_df.columns)
)
aa_df = aaindex_df.dropna(axis=1)
print('\nRevised df ncols:', len(aa_df.columns))
c3 = aa_df.isna().sum()
c4 = c3.sort_values(ascending=False)
print('\nChecking NA in revised df...')
if sum(c4>0):
sys.exit('\nFAIL: aaindex_df still contains cols with NA, please check and drop these before proceeding...')
else:
print('\nPASS: cols with NA successfully dropped from aaindex_df'
, '\nProceeding with combining aa_df with other features_df')
#---------------------------
# aaindex: sanity check 2
#---------------------------
expected_aa_ncols2 = len(aaindex_df.columns) - sum(c2>0)
if len(aa_df.columns) == expected_aa_ncols2:
print('\nPASS: ncols match'
, '\nExpected ncols:', expected_aa_ncols2
, '\nGot:', len(aa_df.columns))
else:
print('\nFAIL: Numbers mismatch'
, '\nExpected ncols:', expected_aa_ncols2
, '\nGot:', len(aa_df.columns))
# Important: need this to identify aaindex cols
aa_df_cols = aa_df.columns
print('\nTotal no. of columns in clean aa_df:', len(aa_df_cols))
###############################################################################
#%% Combining my_features_df and aaindex_df
#===========================
# Merge my_df + aaindex_df
#===========================
if aa_df.columns[aa_df.columns.isin(my_features_df.columns)] == my_features_df.columns[my_features_df.columns.isin(aa_df.columns)]:
print('\nMerging on column: mutationinformation')
if len(my_features_df) == len(aa_df):
expected_nrows = len(my_features_df)
print('\nProceeding to merge, expected nrows in merged_df:', expected_nrows)
else:
sys.exit('\nNrows mismatch, cannot merge. Please check'
, '\nnrows my_df:', len(my_features_df)
, '\nnrows aa_df:', len(aa_df))
#-----------------
# Reset index: mutationinformation
# Very important for merging
#-----------------
aa_df = aa_df.reset_index()
expected_ncols = len(my_features_df.columns) + len(aa_df.columns) - 1 # for the no. of merging col
#-----------------
# Merge: my_features_df + aa_df
#-----------------
merged_df = pd.merge(my_features_df
, aa_df
, on = 'mutationinformation')
#---------------------------
# aaindex: sanity check 3
#---------------------------
if len(merged_df.columns) == expected_ncols:
print('\nPASS: my_features_df and aa_df successfully combined'
, '\nnrows:', len(merged_df)
, '\nncols:', len(merged_df.columns))
else:
sys.exit('\nFAIL: could not combine my_features_df and aa_df'
, '\nCheck dims and merging cols!')
#--------
# Reassign so downstream code doesn't need to change
#--------
my_df = merged_df.copy()
#%% Data: my_df
# Check if non structural pos have crept in
# IDEALLY remove from source! But for rpoB do it here
# Drop NA where numerical cols have them
if gene.lower() in geneL_na_ppi2:
#D1148 get rid of
na_index = my_df['mutationinformation'].index[my_df['mcsm_na_affinity'].apply(np.isnan)]
my_df = my_df.drop(index=na_index)
# FIXED: complete data for all muts inc L114M, F115L, V123L, V125I, V131M
# if gene.lower() in ['embb']:
# na_index = my_df['mutationinformation'].index[my_df['ligand_distance'].apply(np.isnan)]
# my_df = my_df.drop(index=na_index)
# # Sanity check for non-structural positions
# print('\nChecking for non-structural postions')
# na_index = my_df['mutationinformation'].index[my_df['ligand_distance'].apply(np.isnan)]
# if len(na_index) > 0:
# print('\nNon-structural positions detected for gene:', gene.lower()
# , '\nTotal number of these detected:', len(na_index)
# , '\These are at index:', na_index
# , '\nOriginal nrows:', len(my_df)
# , '\nDropping these...')
# my_df = my_df.drop(index=na_index)
# print('\nRevised nrows:', len(my_df))
# else:
# print('\nNo non-structural positions detected for gene:', gene.lower()
# , '\nnrows:', len(my_df))
###########################################################################
#%% Add lineage calculation columns
#FIXME: Check if this can be imported from config?
total_mtblineage_uc = 8
lineage_colnames = ['lineage_list_all', 'lineage_count_all', 'lineage_count_unique', 'lineage_list_unique', 'lineage_multimode']
#bar = my_df[lineage_colnames]
my_df['lineage_proportion'] = my_df['lineage_count_unique']/my_df['lineage_count_all']
my_df['dist_lineage_proportion'] = my_df['lineage_count_unique']/total_mtblineage_uc
###########################################################################
#%% Active site annotation column
# change from numberic to categorical
if my_df['active_site'].dtype in num_type:
my_df['active_site'] = my_df['active_site'].astype(object)
my_df['active_site'].dtype
#%% AA property change
#--------------------
# Water prop change
#--------------------
my_df['water_change'] = my_df['wt_prop_water'] + str('_to_') + my_df['mut_prop_water']
my_df['water_change'].value_counts()
water_prop_changeD = {
'hydrophobic_to_neutral' : 'change'
, 'hydrophobic_to_hydrophobic' : 'no_change'
, 'neutral_to_neutral' : 'no_change'
, 'neutral_to_hydrophobic' : 'change'
, 'hydrophobic_to_hydrophilic' : 'change'
, 'neutral_to_hydrophilic' : 'change'
, 'hydrophilic_to_neutral' : 'change'
, 'hydrophilic_to_hydrophobic' : 'change'
, 'hydrophilic_to_hydrophilic' : 'no_change'
}
my_df['water_change'] = my_df['water_change'].map(water_prop_changeD)
my_df['water_change'].value_counts()
#--------------------
# Polarity change
#--------------------
my_df['polarity_change'] = my_df['wt_prop_polarity'] + str('_to_') + my_df['mut_prop_polarity']
my_df['polarity_change'].value_counts()
polarity_prop_changeD = {
'non-polar_to_non-polar' : 'no_change'
, 'non-polar_to_neutral' : 'change'
, 'neutral_to_non-polar' : 'change'
, 'neutral_to_neutral' : 'no_change'
, 'non-polar_to_basic' : 'change'
, 'acidic_to_neutral' : 'change'
, 'basic_to_neutral' : 'change'
, 'non-polar_to_acidic' : 'change'
, 'neutral_to_basic' : 'change'
, 'acidic_to_non-polar' : 'change'
, 'basic_to_non-polar' : 'change'
, 'neutral_to_acidic' : 'change'
, 'acidic_to_acidic' : 'no_change'
, 'basic_to_acidic' : 'change'
, 'basic_to_basic' : 'no_change'
, 'acidic_to_basic' : 'change'}
my_df['polarity_change'] = my_df['polarity_change'].map(polarity_prop_changeD)
my_df['polarity_change'].value_counts()
#--------------------
# Electrostatics change
#--------------------
my_df['electrostatics_change'] = my_df['wt_calcprop'] + str('_to_') + my_df['mut_calcprop']
my_df['electrostatics_change'].value_counts()
calc_prop_changeD = {
'non-polar_to_non-polar' : 'no_change'
, 'non-polar_to_polar' : 'change'
, 'polar_to_non-polar' : 'change'
, 'non-polar_to_pos' : 'change'
, 'neg_to_non-polar' : 'change'
, 'non-polar_to_neg' : 'change'
, 'pos_to_polar' : 'change'
, 'pos_to_non-polar' : 'change'
, 'polar_to_polar' : 'no_change'
, 'neg_to_neg' : 'no_change'
, 'polar_to_neg' : 'change'
, 'pos_to_neg' : 'change'
, 'pos_to_pos' : 'no_change'
, 'polar_to_pos' : 'change'
, 'neg_to_polar' : 'change'
, 'neg_to_pos' : 'change'
}
my_df['electrostatics_change'] = my_df['electrostatics_change'].map(calc_prop_changeD)
my_df['electrostatics_change'].value_counts()
#--------------------
# Summary change: Create a combined column summarising these three cols
#--------------------
detect_change = 'change'
check_prop_cols = ['water_change', 'polarity_change', 'electrostatics_change']
#my_df['aa_prop_change'] = (my_df.values == detect_change).any(1).astype(int)
my_df['aa_prop_change'] = (my_df[check_prop_cols].values == detect_change).any(1).astype(int)
my_df['aa_prop_change'].value_counts()
my_df['aa_prop_change'].dtype
my_df['aa_prop_change'] = my_df['aa_prop_change'].map({1:'change'
, 0: 'no_change'})
my_df['aa_prop_change'].value_counts()
my_df['aa_prop_change'].dtype
#%% IMPUTE values for OR [check script for exploration: UQ_or_imputer]
#--------------------
# Impute OR values
#--------------------
#or_cols = ['or_mychisq', 'log10_or_mychisq', 'or_fisher']
sel_cols = ['mutationinformation', 'or_mychisq', 'log10_or_mychisq']
or_cols = ['or_mychisq', 'log10_or_mychisq']
print("count of NULL values before imputation\n")
print(my_df[or_cols].isnull().sum())
my_dfI = pd.DataFrame(index = my_df['mutationinformation'] )
my_dfI = pd.DataFrame(KNN(n_neighbors=3, weights="uniform").fit_transform(my_df[or_cols])
, index = my_df['mutationinformation']
, columns = or_cols )
my_dfI.columns = ['or_rawI', 'logorI']
my_dfI.columns
my_dfI = my_dfI.reset_index(drop = False) # prevents old index from being added as a column
my_dfI.head()
print("count of NULL values AFTER imputation\n")
print(my_dfI.isnull().sum())
#-------------------------------------------
# OR df Merge: with original based on index
#-------------------------------------------
#my_df['index_bm'] = my_df.index
mydf_imputed = pd.merge(my_df
, my_dfI
, on = 'mutationinformation')
#mydf_imputed = mydf_imputed.set_index(['index_bm'])
my_df['log10_or_mychisq'].isna().sum()
mydf_imputed['log10_or_mychisq'].isna().sum()
mydf_imputed['logorI'].isna().sum() # should be 0
len(my_df.columns)
len(mydf_imputed.columns)
#-----------------------------------------
# REASSIGN my_df after imputing OR values
#-----------------------------------------
my_df = mydf_imputed.copy()
if my_df['logorI'].isna().sum() == 0:
print('\nPASS: OR values imputed, data ready for ML')
else:
sys.exit('\nFAIL: something went wrong, Data not ready for ML. Please check upstream!')
#!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
#---------------------------------------
# TODO: try other imputation like MICE
#---------------------------------------
#!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
#%%########################################################################
#==========================
# Data for ML
#==========================
my_df_ml = my_df.copy()
# Build column names to mask for affinity chanhes
if gene.lower() in geneL_basic:
#X_stabilityN = common_cols_stabiltyN
gene_affinity_colnames = []# not needed as its the common ones
cols_to_mask = ['ligand_affinity_change']
if gene.lower() in geneL_ppi2:
gene_affinity_colnames = ['mcsm_ppi2_affinity', 'interface_dist']
#X_stabilityN = common_cols_stabiltyN + geneL_ppi2_st_cols
cols_to_mask = ['ligand_affinity_change', 'mcsm_ppi2_affinity']
if gene.lower() in geneL_na:
gene_affinity_colnames = ['mcsm_na_affinity']
#X_stabilityN = common_cols_stabiltyN + geneL_na_st_cols
cols_to_mask = ['ligand_affinity_change', 'mcsm_na_affinity']
if gene.lower() in geneL_na_ppi2:
gene_affinity_colnames = ['mcsm_na_affinity'] + ['mcsm_ppi2_affinity', 'interface_dist']
#X_stabilityN = common_cols_stabiltyN + geneL_na_ppi2_st_cols
cols_to_mask = ['ligand_affinity_change', 'mcsm_na_affinity', 'mcsm_ppi2_affinity']
#=======================
# Masking columns:
# (mCSM-lig, mCSM-NA, mCSM-ppi2) values for lig_dist >10
#=======================
my_df_ml['mutationinformation'][my_df_ml['ligand_distance']>10].value_counts()
my_df_ml.groupby('mutationinformation')['ligand_distance'].apply(lambda x: (x>10)).value_counts()
my_df_ml.loc[(my_df_ml['ligand_distance'] > 10), cols_to_mask].value_counts()
# mask the mcsm affinity related columns where ligand distance > 10
my_df_ml.loc[(my_df_ml['ligand_distance'] > 10), cols_to_mask] = 0
(my_df_ml['ligand_affinity_change'] == 0).sum()
mask_check = my_df_ml[['mutationinformation', 'ligand_distance'] + cols_to_mask]
#===================================================
# write file for check
mask_check.sort_values(by = ['ligand_distance'], ascending = True, inplace = True)
mask_check.to_csv(outdir + 'ml/' + gene.lower() + '_mask_check.csv')
#===================================================
###############################################################################
#%% Feature groups (FG): Build X for Input ML
############################################################################
#===========================
# FG1: Evolutionary features
#===========================
X_evolFN = ['consurf_score'
, 'snap2_score'
, 'provean_score']
###############################################################################
#========================
# FG2: Stability features
#========================
#--------
# common
#--------
X_common_stability_Fnum = [
'duet_stability_change'
, 'ddg_foldx'
, 'deepddg'
, 'ddg_dynamut2'
, 'contacts']
#--------
# FoldX
#--------
X_foldX_Fnum = [ 'electro_rr', 'electro_mm', 'electro_sm', 'electro_ss'
, 'disulfide_rr', 'disulfide_mm', 'disulfide_sm', 'disulfide_ss'
, 'hbonds_rr', 'hbonds_mm', 'hbonds_sm', 'hbonds_ss'
, 'partcov_rr', 'partcov_mm', 'partcov_sm', 'partcov_ss'
, 'vdwclashes_rr', 'vdwclashes_mm', 'vdwclashes_sm', 'vdwclashes_ss'
, 'volumetric_rr', 'volumetric_mm', 'volumetric_ss']
X_stability_FN = X_common_stability_Fnum + X_foldX_Fnum
###############################################################################
#===================
# FG3: Affinity features
#===================
common_affinity_Fnum = ['ligand_distance'
, 'ligand_affinity_change'
, 'mmcsm_lig']
# if gene.lower() in geneL_basic:
# X_affinityFN = common_affinity_Fnum
# else:
# X_affinityFN = common_affinity_Fnum + gene_affinity_colnames
X_affinityFN = common_affinity_Fnum + gene_affinity_colnames
###############################################################################
#============================
# FG4: Residue level features
#============================
#-----------
# AA index
#-----------
X_aaindex_Fnum = list(aa_df_cols)
print('\nTotal no. of features for aaindex:', len(X_aaindex_Fnum))
#-----------------
# surface area
# depth
# hydrophobicity
#-----------------
X_str_Fnum = ['rsa'
#, 'asa'
, 'kd_values'
, 'rd_values']
#---------------------------
# Other aa properties
# active site indication
#---------------------------
X_aap_Fcat = ['ss_class'
# , 'wt_prop_water'
# , 'mut_prop_water'
# , 'wt_prop_polarity'
# , 'mut_prop_polarity'
# , 'wt_calcprop'
# , 'mut_calcprop'
, 'aa_prop_change'
, 'electrostatics_change'
, 'polarity_change'
, 'water_change'
, 'active_site']
X_resprop_FN = X_aaindex_Fnum + X_str_Fnum + X_aap_Fcat
###############################################################################
#========================
# FG5: Genomic features
#========================
X_gn_mafor_Fnum = ['maf'
#, 'logorI'
# , 'or_rawI'
# , 'or_mychisq'
# , 'or_logistic'
# , 'or_fisher'
# , 'pval_fisher'
]
X_gn_linegae_Fnum = ['lineage_proportion'
, 'dist_lineage_proportion'
#, 'lineage' # could be included as a category but it has L2;L4 formatting
, 'lineage_count_all'
, 'lineage_count_unique'
]
# X_gn_Fcat = ['drtype_mode_labels' # beware then you can't use it to predict [USED it for uq_v1, not v2]
# #, 'gene_name' # will be required for the combined stuff
# ]
X_gn_Fcat = []
X_genomicFN = X_gn_mafor_Fnum + X_gn_linegae_Fnum + X_gn_Fcat
###############################################################################
#========================
# FG6 collapsed: Structural : Atability + Affinity + ResidueProp
#========================
X_structural_FN = X_stability_FN + X_affinityFN + X_resprop_FN
###############################################################################
#========================
# BUILDING all features
#========================
all_featuresN = X_evolFN + X_structural_FN + X_genomicFN
###############################################################################
#%% Define training and test data
#================================================================
# Training and BLIND test set: 80/20
# dst with actual values : training set
# dst with imputed values : THROW AWAY [unrepresentative]
#================================================================
my_df_ml[drug].isna().sum()
# blind_test_df = my_df_ml[my_df_ml[drug].isna()]
# blind_test_df.shape
#training_df = my_df_ml[my_df_ml[drug].notna()]
#training_df.shape
training_df = my_df_ml.copy()
# Target 1: dst_mode
training_df[drug].value_counts()
training_df['dst_mode'].value_counts()
####################################################################
#====================================
# ML data: Train test split: 80/20
# with stratification
# 80% : training_data for CV
# 20% : blind test
#=====================================
x_features = training_df[all_featuresN]
y_target = training_df['dst_mode']
# sanity check
if not 'dst_mode' in x_features.columns:
print('\nPASS: x_features has no target variable')
x_ncols = len(x_features.columns)
print('\nNo. of columns for x_features:', x_ncols)
# NEED It for scaling law split
#https://towardsdatascience.com/finally-why-we-use-an-80-20-split-for-training-and-test-data-plus-an-alternative-method-oh-yes-edc77e96295d
else:
sys.exit('\nFAIL: x_features has target variable included. FIX it and rerun!')
#-------------------
# train-test split
#-------------------
#x_train, x_test, y_train, y_test # traditional var_names
# so my downstream code doesn't need to change
X, X_bts, y, y_bts = train_test_split(x_features, y_target
, test_size = 0.2
, **rs
, stratify = y_target)
yc1 = Counter(y)
yc1_ratio = yc1[0]/yc1[1]
yc2 = Counter(y_bts)
yc2_ratio = yc2[0]/yc2[1]
###############################################################################
#======================================================
# Determine categorical and numerical features
#======================================================
numerical_cols = X.select_dtypes(include=['int64', 'float64']).columns
numerical_cols
categorical_cols = X.select_dtypes(include=['object', 'bool']).columns
categorical_cols
################################################################################
# IMPORTANT sanity checks
if len(X.columns) == len(X_evolFN) + len(X_stability_FN) + len(X_affinityFN) + len(X_resprop_FN) + len(X_genomicFN):
print('\nPASS: ML data with input features, training and test generated...'
, '\n\nTotal no. of input features:' , len(X.columns)
, '\n--------No. of numerical features:' , len(numerical_cols)
, '\n--------No. of categorical features:' , len(categorical_cols)
, '\n\nTotal no. of evolutionary features:' , len(X_evolFN)
, '\n\nTotal no. of stability features:' , len(X_stability_FN)
, '\n--------Common stabilty cols:' , len(X_common_stability_Fnum)
, '\n--------Foldx cols:' , len(X_foldX_Fnum)
, '\n\nTotal no. of affinity features:' , len(X_affinityFN)
, '\n--------Common affinity cols:' , len(common_affinity_Fnum)
, '\n--------Gene specific affinity cols:' , len(gene_affinity_colnames)
, '\n\nTotal no. of residue level features:', len(X_resprop_FN)
, '\n--------AA index cols:' , len(X_aaindex_Fnum)
, '\n--------Residue Prop cols:' , len(X_str_Fnum)
, '\n--------AA change Prop cols:' , len(X_aap_Fcat)
, '\n\nTotal no. of genomic features:' , len(X_genomicFN)
, '\n--------MAF+OR cols:' , len(X_gn_mafor_Fnum)
, '\n--------Lineage cols:' , len(X_gn_linegae_Fnum)
, '\n--------Other cols:' , len(X_gn_Fcat)
)
else:
print('\nFAIL: numbers mismatch'
, '\nExpected:',len(X_evolFN) + len(X_stability_FN) + len(X_affinityFN) + len(X_resprop_FN) + len(X_genomicFN)
, '\nGot:', len(X.columns))
sys.exit()
###############################################################################
print('\n-------------------------------------------------------------'
, '\nSuccessfully split data: ALL features'
, '\nactual values: training set'
, '\nSplit:', tts_split
#, '\nimputed values: blind test set'
, '\n\nTotal data size:', len(X) + len(X_bts)
, '\n\nTrain data size:', X.shape
, '\ny_train numbers:', yc1
, '\n\nTest data size:', X_bts.shape
, '\ny_test_numbers:', yc2
, '\n\ny_train ratio:',yc1_ratio
, '\ny_test ratio:', yc2_ratio
, '\n-------------------------------------------------------------'
)
##########################################################################
# Quick check
#(X['ligand_affinity_change']==0).sum() == (X['ligand_distance']>10).sum()
for i in range(len(cols_to_mask)):
ind = i+1
print('\nindex:', i, '\nind:', ind)
print('\nMask count check:'
, (my_df_ml[cols_to_mask[i]]==0).sum() == (my_df_ml['ligand_distance']>10).sum()
)
print('Original Data\n', Counter(y)
, 'Data dim:', X.shape)
###########################################################################
#%%
###########################################################################
# RESAMPLING
###########################################################################
#------------------------------
# Simple Random oversampling
# [Numerical + catgeorical]
#------------------------------
oversample = RandomOverSampler(sampling_strategy='minority')
X_ros, y_ros = oversample.fit_resample(X, y)
print('\nSimple Random OverSampling\n', Counter(y_ros))
print(X_ros.shape)
#------------------------------
# Simple Random Undersampling
# [Numerical + catgeorical]
#------------------------------
undersample = RandomUnderSampler(sampling_strategy='majority')
X_rus, y_rus = undersample.fit_resample(X, y)
print('\nSimple Random UnderSampling\n', Counter(y_rus))
print(X_rus.shape)
#------------------------------
# Simple combine ROS and RUS
# [Numerical + catgeorical]
#------------------------------
oversample = RandomOverSampler(sampling_strategy='minority')
X_ros, y_ros = oversample.fit_resample(X, y)
undersample = RandomUnderSampler(sampling_strategy='majority')
X_rouC, y_rouC = undersample.fit_resample(X_ros, y_ros)
print('\nSimple Combined Over and UnderSampling\n', Counter(y_rouC))
print(X_rouC.shape)
#------------------------------
# SMOTE_NC: oversampling
# [numerical + categorical]
#https://stackoverflow.com/questions/47655813/oversampling-smote-for-binary-and-categorical-data-in-python
#------------------------------
# Determine categorical and numerical features
numerical_ix = X.select_dtypes(include=['int64', 'float64']).columns
numerical_ix
num_featuresL = list(numerical_ix)
numerical_colind = X.columns.get_indexer(list(numerical_ix) )
numerical_colind
categorical_ix = X.select_dtypes(include=['object', 'bool']).columns
categorical_ix
categorical_colind = X.columns.get_indexer(list(categorical_ix))
categorical_colind
k_sm = 5 # 5 is default
sm_nc = SMOTENC(categorical_features=categorical_colind, k_neighbors = k_sm, **rs, **njobs)
X_smnc, y_smnc = sm_nc.fit_resample(X, y)
print('\nSMOTE_NC OverSampling\n', Counter(y_smnc))
print(X_smnc.shape)
globals().update(locals()) # TROLOLOLOLOLOLS
#print("i did a horrible hack :-)")
###############################################################################
#%% SMOTE RESAMPLING for NUMERICAL ONLY*
# #------------------------------
# # SMOTE: Oversampling
# # [Numerical ONLY]
# #------------------------------
# k_sm = 1
# sm = SMOTE(sampling_strategy = 'auto', k_neighbors = k_sm, **rs)
# X_sm, y_sm = sm.fit_resample(X, y)
# print(X_sm.shape)
# print('\nSMOTE OverSampling\n', Counter(y_sm))
# y_sm_df = y_sm.to_frame()
# y_sm_df.value_counts().plot(kind = 'bar')
# #------------------------------
# # SMOTE: Over + Undersampling COMBINED
# # [Numerical ONLY]
# #-----------------------------
# sm_enn = SMOTEENN(enn=EditedNearestNeighbours(sampling_strategy='all', **rs, **njobs ))
# X_enn, y_enn = sm_enn.fit_resample(X, y)
# print(X_enn.shape)
# print('\nSMOTE Over+Under Sampling combined\n', Counter(y_enn))
###############################################################################
# TODO: Find over and undersampling JUST for categorical data
###########################################################################
print('\n#################################################################'
, '\nDim of X for gene:', gene.lower(), '\n', X.shape
, '\n###############################################################')

814
scripts/ml/ml_data_cd_sl.py
View file

@ -1,814 +0,0 @@
#!/usr/bin/env python3
# -*- coding: utf-8 -*-
"""
Created on Sun Mar 6 13:41:54 2022
@author: tanu
"""
def setvars(gene,drug):
#https://stackoverflow.com/questions/51695322/compare-multiple-algorithms-with-sklearn-pipeline
import os, sys
import pandas as pd
import numpy as np
print(np.__version__)
print(pd.__version__)
import pprint as pp
from copy import deepcopy
from collections import Counter
from sklearn.impute import KNNImputer as KNN
from imblearn.over_sampling import RandomOverSampler
from imblearn.under_sampling import RandomUnderSampler
from imblearn.over_sampling import SMOTE
from sklearn.datasets import make_classification
from imblearn.combine import SMOTEENN
from imblearn.combine import SMOTETomek
from imblearn.over_sampling import SMOTENC
from imblearn.under_sampling import EditedNearestNeighbours
from imblearn.under_sampling import RepeatedEditedNearestNeighbours
from sklearn.metrics import make_scorer, confusion_matrix, accuracy_score, balanced_accuracy_score, precision_score, average_precision_score, recall_score
from sklearn.metrics import roc_auc_score, roc_curve, f1_score, matthews_corrcoef, jaccard_score, classification_report
from sklearn.model_selection import train_test_split, cross_validate, cross_val_score
from sklearn.model_selection import StratifiedKFold,RepeatedStratifiedKFold, RepeatedKFold
from sklearn.pipeline import Pipeline, make_pipeline
import argparse
import re
#%% GLOBALS
tts_split = "sl"
rs = {'random_state': 42}
njobs = {'n_jobs': 10}
scoring_fn = ({ 'mcc' : make_scorer(matthews_corrcoef)
, 'accuracy' : make_scorer(accuracy_score)
, 'fscore' : make_scorer(f1_score)
, 'precision' : make_scorer(precision_score)
, 'recall' : make_scorer(recall_score)
, 'roc_auc' : make_scorer(roc_auc_score)
, 'jcc' : make_scorer(jaccard_score)
})
skf_cv = StratifiedKFold(n_splits = 10
#, shuffle = False, random_state= None)
, shuffle = True,**rs)
rskf_cv = RepeatedStratifiedKFold(n_splits = 10
, n_repeats = 3
, **rs)
mcc_score_fn = {'mcc': make_scorer(matthews_corrcoef)}
jacc_score_fn = {'jcc': make_scorer(jaccard_score)}
#%% FOR LATER: Combine ED logo data
###########################################################################
homedir = os.path.expanduser("~")
geneL_basic = ['pnca']
geneL_na = ['gid']
geneL_na_ppi2 = ['rpob']
geneL_ppi2 = ['alr', 'embb', 'katg']
#num_type = ['int64', 'float64']
num_type = ['int16', 'int32', 'int64', 'float16', 'float32', 'float64']
cat_type = ['object', 'bool']
#==============
# directories
#==============
datadir = homedir + '/git/Data/'
indir = datadir + drug + '/input/'
outdir = datadir + drug + '/output/'
#=======
# input
#=======
#---------
# File 1
#---------
infile_ml1 = outdir + gene.lower() + '_merged_df3.csv'
#infile_ml2 = outdir + gene.lower() + '_merged_df2.csv'
my_features_df = pd.read_csv(infile_ml1, index_col = 0)
my_features_df = my_features_df .reset_index(drop = True)
my_features_df.index
my_features_df.dtypes
mycols = my_features_df.columns
#---------
# File 2
#---------
infile_aaindex = outdir + 'aa_index/' + gene.lower() + '_aa.csv'
aaindex_df = pd.read_csv(infile_aaindex, index_col = 0)
aaindex_df.dtypes
#-----------
# check for non-numerical columns
#-----------
if any(aaindex_df.dtypes==object):
print('\naaindex_df contains non-numerical data')
aaindex_df_object = aaindex_df.select_dtypes(include = cat_type)
print('\nTotal no. of non-numerial columns:', len(aaindex_df_object.columns))
expected_aa_ncols = len(aaindex_df.columns) - len(aaindex_df_object.columns)
#-----------
# Extract numerical data only
#-----------
print('\nSelecting numerical data only')
aaindex_df = aaindex_df.select_dtypes(include = num_type)
#---------------------------
# aaindex: sanity check 1
#---------------------------
if len(aaindex_df.columns) == expected_aa_ncols:
print('\nPASS: successfully selected numerical columns only for aaindex_df')
else:
print('\nFAIL: Numbers mismatch'
, '\nExpected ncols:', expected_aa_ncols
, '\nGot:', len(aaindex_df.columns))
#---------------
# check for NA
#---------------
print('\nNow checking for NA in the remaining aaindex_cols')
c1 = aaindex_df.isna().sum()
c2 = c1.sort_values(ascending=False)
print('\nCounting aaindex_df cols with NA'
, '\nncols with NA:', sum(c2>0), 'columns'
, '\nDropping these...'
, '\nOriginal ncols:', len(aaindex_df.columns)
)
aa_df = aaindex_df.dropna(axis=1)
print('\nRevised df ncols:', len(aa_df.columns))
c3 = aa_df.isna().sum()
c4 = c3.sort_values(ascending=False)
print('\nChecking NA in revised df...')
if sum(c4>0):
sys.exit('\nFAIL: aaindex_df still contains cols with NA, please check and drop these before proceeding...')
else:
print('\nPASS: cols with NA successfully dropped from aaindex_df'
, '\nProceeding with combining aa_df with other features_df')
#---------------------------
# aaindex: sanity check 2
#---------------------------
expected_aa_ncols2 = len(aaindex_df.columns) - sum(c2>0)
if len(aa_df.columns) == expected_aa_ncols2:
print('\nPASS: ncols match'
, '\nExpected ncols:', expected_aa_ncols2
, '\nGot:', len(aa_df.columns))
else:
print('\nFAIL: Numbers mismatch'
, '\nExpected ncols:', expected_aa_ncols2
, '\nGot:', len(aa_df.columns))
# Important: need this to identify aaindex cols
aa_df_cols = aa_df.columns
print('\nTotal no. of columns in clean aa_df:', len(aa_df_cols))
###############################################################################
#%% Combining my_features_df and aaindex_df
#===========================
# Merge my_df + aaindex_df
#===========================
if aa_df.columns[aa_df.columns.isin(my_features_df.columns)] == my_features_df.columns[my_features_df.columns.isin(aa_df.columns)]:
print('\nMerging on column: mutationinformation')
if len(my_features_df) == len(aa_df):
expected_nrows = len(my_features_df)
print('\nProceeding to merge, expected nrows in merged_df:', expected_nrows)
else:
sys.exit('\nNrows mismatch, cannot merge. Please check'
, '\nnrows my_df:', len(my_features_df)
, '\nnrows aa_df:', len(aa_df))
#-----------------
# Reset index: mutationinformation
# Very important for merging
#-----------------
aa_df = aa_df.reset_index()
expected_ncols = len(my_features_df.columns) + len(aa_df.columns) - 1 # for the no. of merging col
#-----------------
# Merge: my_features_df + aa_df
#-----------------
merged_df = pd.merge(my_features_df
, aa_df
, on = 'mutationinformation')
#---------------------------
# aaindex: sanity check 3
#---------------------------
if len(merged_df.columns) == expected_ncols:
print('\nPASS: my_features_df and aa_df successfully combined'
, '\nnrows:', len(merged_df)
, '\nncols:', len(merged_df.columns))
else:
sys.exit('\nFAIL: could not combine my_features_df and aa_df'
, '\nCheck dims and merging cols!')
#--------
# Reassign so downstream code doesn't need to change
#--------
my_df = merged_df.copy()
#%% Data: my_df
# Check if non structural pos have crept in
# IDEALLY remove from source! But for rpoB do it here
# Drop NA where numerical cols have them
if gene.lower() in geneL_na_ppi2:
#D1148 get rid of
na_index = my_df['mutationinformation'].index[my_df['mcsm_na_affinity'].apply(np.isnan)]
my_df = my_df.drop(index=na_index)
# FIXED: complete data for all muts inc L114M, F115L, V123L, V125I, V131M
# if gene.lower() in ['embb']:
# na_index = my_df['mutationinformation'].index[my_df['ligand_distance'].apply(np.isnan)]
# my_df = my_df.drop(index=na_index)
# # Sanity check for non-structural positions
# print('\nChecking for non-structural postions')
# na_index = my_df['mutationinformation'].index[my_df['ligand_distance'].apply(np.isnan)]
# if len(na_index) > 0:
# print('\nNon-structural positions detected for gene:', gene.lower()
# , '\nTotal number of these detected:', len(na_index)
# , '\These are at index:', na_index
# , '\nOriginal nrows:', len(my_df)
# , '\nDropping these...')
# my_df = my_df.drop(index=na_index)
# print('\nRevised nrows:', len(my_df))
# else:
# print('\nNo non-structural positions detected for gene:', gene.lower()
# , '\nnrows:', len(my_df))
###########################################################################
#%% Add lineage calculation columns
#FIXME: Check if this can be imported from config?
total_mtblineage_uc = 8
lineage_colnames = ['lineage_list_all', 'lineage_count_all', 'lineage_count_unique', 'lineage_list_unique', 'lineage_multimode']
#bar = my_df[lineage_colnames]
my_df['lineage_proportion'] = my_df['lineage_count_unique']/my_df['lineage_count_all']
my_df['dist_lineage_proportion'] = my_df['lineage_count_unique']/total_mtblineage_uc
###########################################################################
#%% Active site annotation column
# change from numberic to categorical
if my_df['active_site'].dtype in num_type:
my_df['active_site'] = my_df['active_site'].astype(object)
my_df['active_site'].dtype
#%% AA property change
#--------------------
# Water prop change
#--------------------
my_df['water_change'] = my_df['wt_prop_water'] + str('_to_') + my_df['mut_prop_water']
my_df['water_change'].value_counts()
water_prop_changeD = {
'hydrophobic_to_neutral' : 'change'
, 'hydrophobic_to_hydrophobic' : 'no_change'
, 'neutral_to_neutral' : 'no_change'
, 'neutral_to_hydrophobic' : 'change'
, 'hydrophobic_to_hydrophilic' : 'change'
, 'neutral_to_hydrophilic' : 'change'
, 'hydrophilic_to_neutral' : 'change'
, 'hydrophilic_to_hydrophobic' : 'change'
, 'hydrophilic_to_hydrophilic' : 'no_change'
}
my_df['water_change'] = my_df['water_change'].map(water_prop_changeD)
my_df['water_change'].value_counts()
#--------------------
# Polarity change
#--------------------
my_df['polarity_change'] = my_df['wt_prop_polarity'] + str('_to_') + my_df['mut_prop_polarity']
my_df['polarity_change'].value_counts()
polarity_prop_changeD = {
'non-polar_to_non-polar' : 'no_change'
, 'non-polar_to_neutral' : 'change'
, 'neutral_to_non-polar' : 'change'
, 'neutral_to_neutral' : 'no_change'
, 'non-polar_to_basic' : 'change'
, 'acidic_to_neutral' : 'change'
, 'basic_to_neutral' : 'change'
, 'non-polar_to_acidic' : 'change'
, 'neutral_to_basic' : 'change'
, 'acidic_to_non-polar' : 'change'
, 'basic_to_non-polar' : 'change'
, 'neutral_to_acidic' : 'change'
, 'acidic_to_acidic' : 'no_change'
, 'basic_to_acidic' : 'change'
, 'basic_to_basic' : 'no_change'
, 'acidic_to_basic' : 'change'}
my_df['polarity_change'] = my_df['polarity_change'].map(polarity_prop_changeD)
my_df['polarity_change'].value_counts()
#--------------------
# Electrostatics change
#--------------------
my_df['electrostatics_change'] = my_df['wt_calcprop'] + str('_to_') + my_df['mut_calcprop']
my_df['electrostatics_change'].value_counts()
calc_prop_changeD = {
'non-polar_to_non-polar' : 'no_change'
, 'non-polar_to_polar' : 'change'
, 'polar_to_non-polar' : 'change'
, 'non-polar_to_pos' : 'change'
, 'neg_to_non-polar' : 'change'
, 'non-polar_to_neg' : 'change'
, 'pos_to_polar' : 'change'
, 'pos_to_non-polar' : 'change'
, 'polar_to_polar' : 'no_change'
, 'neg_to_neg' : 'no_change'
, 'polar_to_neg' : 'change'
, 'pos_to_neg' : 'change'
, 'pos_to_pos' : 'no_change'
, 'polar_to_pos' : 'change'
, 'neg_to_polar' : 'change'
, 'neg_to_pos' : 'change'
}
my_df['electrostatics_change'] = my_df['electrostatics_change'].map(calc_prop_changeD)
my_df['electrostatics_change'].value_counts()
#--------------------
# Summary change: Create a combined column summarising these three cols
#--------------------
detect_change = 'change'
check_prop_cols = ['water_change', 'polarity_change', 'electrostatics_change']
#my_df['aa_prop_change'] = (my_df.values == detect_change).any(1).astype(int)
my_df['aa_prop_change'] = (my_df[check_prop_cols].values == detect_change).any(1).astype(int)
my_df['aa_prop_change'].value_counts()
my_df['aa_prop_change'].dtype
my_df['aa_prop_change'] = my_df['aa_prop_change'].map({1:'change'
, 0: 'no_change'})
my_df['aa_prop_change'].value_counts()
my_df['aa_prop_change'].dtype
#%% IMPUTE values for OR [check script for exploration: UQ_or_imputer]
#--------------------
# Impute OR values
#--------------------
#or_cols = ['or_mychisq', 'log10_or_mychisq', 'or_fisher']
sel_cols = ['mutationinformation', 'or_mychisq', 'log10_or_mychisq']
or_cols = ['or_mychisq', 'log10_or_mychisq']
print("count of NULL values before imputation\n")
print(my_df[or_cols].isnull().sum())
my_dfI = pd.DataFrame(index = my_df['mutationinformation'] )
my_dfI = pd.DataFrame(KNN(n_neighbors=3, weights="uniform").fit_transform(my_df[or_cols])
, index = my_df['mutationinformation']
, columns = or_cols )
my_dfI.columns = ['or_rawI', 'logorI']
my_dfI.columns
my_dfI = my_dfI.reset_index(drop = False) # prevents old index from being added as a column
my_dfI.head()
print("count of NULL values AFTER imputation\n")
print(my_dfI.isnull().sum())
#-------------------------------------------
# OR df Merge: with original based on index
#-------------------------------------------
#my_df['index_bm'] = my_df.index
mydf_imputed = pd.merge(my_df
, my_dfI
, on = 'mutationinformation')
#mydf_imputed = mydf_imputed.set_index(['index_bm'])
my_df['log10_or_mychisq'].isna().sum()
mydf_imputed['log10_or_mychisq'].isna().sum()
mydf_imputed['logorI'].isna().sum() # should be 0
len(my_df.columns)
len(mydf_imputed.columns)
#-----------------------------------------
# REASSIGN my_df after imputing OR values
#-----------------------------------------
my_df = mydf_imputed.copy()
if my_df['logorI'].isna().sum() == 0:
print('\nPASS: OR values imputed, data ready for ML')
else:
sys.exit('\nFAIL: something went wrong, Data not ready for ML. Please check upstream!')
#!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
#---------------------------------------
# TODO: try other imputation like MICE
#---------------------------------------
#!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
#%%########################################################################
#==========================
# Data for ML
#==========================
my_df_ml = my_df.copy()
# Build column names to mask for affinity chanhes
if gene.lower() in geneL_basic:
#X_stabilityN = common_cols_stabiltyN
gene_affinity_colnames = []# not needed as its the common ones
cols_to_mask = ['ligand_affinity_change']
if gene.lower() in geneL_ppi2:
gene_affinity_colnames = ['mcsm_ppi2_affinity', 'interface_dist']
#X_stabilityN = common_cols_stabiltyN + geneL_ppi2_st_cols
cols_to_mask = ['ligand_affinity_change', 'mcsm_ppi2_affinity']
if gene.lower() in geneL_na:
gene_affinity_colnames = ['mcsm_na_affinity']
#X_stabilityN = common_cols_stabiltyN + geneL_na_st_cols
cols_to_mask = ['ligand_affinity_change', 'mcsm_na_affinity']
if gene.lower() in geneL_na_ppi2:
gene_affinity_colnames = ['mcsm_na_affinity'] + ['mcsm_ppi2_affinity', 'interface_dist']
#X_stabilityN = common_cols_stabiltyN + geneL_na_ppi2_st_cols
cols_to_mask = ['ligand_affinity_change', 'mcsm_na_affinity', 'mcsm_ppi2_affinity']
#=======================
# Masking columns:
# (mCSM-lig, mCSM-NA, mCSM-ppi2) values for lig_dist >10
#=======================
my_df_ml['mutationinformation'][my_df_ml['ligand_distance']>10].value_counts()
my_df_ml.groupby('mutationinformation')['ligand_distance'].apply(lambda x: (x>10)).value_counts()
my_df_ml.loc[(my_df_ml['ligand_distance'] > 10), cols_to_mask].value_counts()
# mask the mcsm affinity related columns where ligand distance > 10
my_df_ml.loc[(my_df_ml['ligand_distance'] > 10), cols_to_mask] = 0
(my_df_ml['ligand_affinity_change'] == 0).sum()
mask_check = my_df_ml[['mutationinformation', 'ligand_distance'] + cols_to_mask]
#===================================================
# write file for check
mask_check.sort_values(by = ['ligand_distance'], ascending = True, inplace = True)
mask_check.to_csv(outdir + 'ml/' + gene.lower() + '_mask_check.csv')
#===================================================
###############################################################################
#%% Feature groups (FG): Build X for Input ML
############################################################################
#===========================
# FG1: Evolutionary features
#===========================
X_evolFN = ['consurf_score'
, 'snap2_score'
, 'provean_score']
###############################################################################
#========================
# FG2: Stability features
#========================
#--------
# common
#--------
X_common_stability_Fnum = [
'duet_stability_change'
, 'ddg_foldx'
, 'deepddg'
, 'ddg_dynamut2'
, 'contacts']
#--------
# FoldX
#--------
X_foldX_Fnum = [ 'electro_rr', 'electro_mm', 'electro_sm', 'electro_ss'
, 'disulfide_rr', 'disulfide_mm', 'disulfide_sm', 'disulfide_ss'
, 'hbonds_rr', 'hbonds_mm', 'hbonds_sm', 'hbonds_ss'
, 'partcov_rr', 'partcov_mm', 'partcov_sm', 'partcov_ss'
, 'vdwclashes_rr', 'vdwclashes_mm', 'vdwclashes_sm', 'vdwclashes_ss'
, 'volumetric_rr', 'volumetric_mm', 'volumetric_ss']
X_stability_FN = X_common_stability_Fnum + X_foldX_Fnum
###############################################################################
#===================
# FG3: Affinity features
#===================
common_affinity_Fnum = ['ligand_distance'
, 'ligand_affinity_change'
, 'mmcsm_lig']
# if gene.lower() in geneL_basic:
# X_affinityFN = common_affinity_Fnum
# else:
# X_affinityFN = common_affinity_Fnum + gene_affinity_colnames
X_affinityFN = common_affinity_Fnum + gene_affinity_colnames
###############################################################################
#============================
# FG4: Residue level features
#============================
#-----------
# AA index
#-----------
X_aaindex_Fnum = list(aa_df_cols)
print('\nTotal no. of features for aaindex:', len(X_aaindex_Fnum))
#-----------------
# surface area
# depth
# hydrophobicity
#-----------------
X_str_Fnum = ['rsa'
#, 'asa'
, 'kd_values'
, 'rd_values']
#---------------------------
# Other aa properties
# active site indication
#---------------------------
X_aap_Fcat = ['ss_class'
# , 'wt_prop_water'
# , 'mut_prop_water'
# , 'wt_prop_polarity'
# , 'mut_prop_polarity'
# , 'wt_calcprop'
# , 'mut_calcprop'
, 'aa_prop_change'
, 'electrostatics_change'
, 'polarity_change'
, 'water_change'
, 'active_site']
X_resprop_FN = X_aaindex_Fnum + X_str_Fnum + X_aap_Fcat
###############################################################################
#========================
# FG5: Genomic features
#========================
X_gn_mafor_Fnum = ['maf'
#, 'logorI'
# , 'or_rawI'
# , 'or_mychisq'
# , 'or_logistic'
# , 'or_fisher'
# , 'pval_fisher'
]
X_gn_linegae_Fnum = ['lineage_proportion'
, 'dist_lineage_proportion'
#, 'lineage' # could be included as a category but it has L2;L4 formatting
, 'lineage_count_all'
, 'lineage_count_unique'
]
# X_gn_Fcat = ['drtype_mode_labels' # beware then you can't use it to predict [USED it for uq_v1, not v2]
# #, 'gene_name' # will be required for the combined stuff
# ]
X_gn_Fcat = []
X_genomicFN = X_gn_mafor_Fnum + X_gn_linegae_Fnum + X_gn_Fcat
###############################################################################
#========================
# FG6 collapsed: Structural : Atability + Affinity + ResidueProp
#========================
X_structural_FN = X_stability_FN + X_affinityFN + X_resprop_FN
###############################################################################
#========================
# BUILDING all features
#========================
all_featuresN = X_evolFN + X_structural_FN + X_genomicFN
###############################################################################
#%% Define training and test data
#================================================================
# Training and BLIND test set: scaling law split
#https://towardsdatascience.com/finally-why-we-use-an-80-20-split-for-training-and-test-data-plus-an-alternative-method-oh-yes-edc77e96295d
# dst with actual values : training set
# dst with imputed values : THROW AWAY [unrepresentative]
# test data size ~ 1/sqrt(features NOT including target variable)
#================================================================
my_df_ml[drug].isna().sum()
# blind_test_df = my_df_ml[my_df_ml[drug].isna()]
# blind_test_df.shape
#training_df = my_df_ml[my_df_ml[drug].notna()]
#training_df.shape
training_df = my_df_ml.copy()
# Target 1: dst_mode
training_df[drug].value_counts()
training_df['dst_mode'].value_counts()
####################################################################
#====================================
# ML data: Train test split: SL
# with stratification
# 1-blind test : training_data for CV
# 1/sqrt(columns) : blind test
#===========================================
x_features = training_df[all_featuresN]
y_target = training_df['dst_mode']
# sanity check
if not 'dst_mode' in x_features.columns:
print('\nPASS: x_features has no target variable')
x_ncols = len(x_features.columns)
print('\nNo. of columns for x_features:', x_ncols)
# NEED It for scaling law split
#https://towardsdatascience.com/finally-why-we-use-an-80-20-split-for-training-and-test-data-plus-an-alternative-method-oh-yes-edc77e96295d
else:
sys.exit('\nFAIL: x_features has target variable included. FIX it and rerun!')
#-------------------
# train-test split
#-------------------
sl_test_size = 1/np.sqrt(x_ncols)
train = 1 - sl_test_size
#x_train, x_test, y_train, y_test # traditional var_names
# so my downstream code doesn't need to change
X, X_bts, y, y_bts = train_test_split(x_features, y_target
, test_size = sl_test_size
, **rs
, stratify = y_target)
yc1 = Counter(y)
yc1_ratio = yc1[0]/yc1[1]
yc2 = Counter(y_bts)
yc2_ratio = yc2[0]/yc2[1]
###############################################################################
#======================================================
# Determine categorical and numerical features
#======================================================
numerical_cols = X.select_dtypes(include=['int64', 'float64']).columns
numerical_cols
categorical_cols = X.select_dtypes(include=['object', 'bool']).columns
categorical_cols
################################################################################
# IMPORTANT sanity checks
if len(X.columns) == len(X_evolFN) + len(X_stability_FN) + len(X_affinityFN) + len(X_resprop_FN) + len(X_genomicFN):
print('\nPASS: ML data with input features, training and test generated...'
, '\n\nTotal no. of input features:' , len(X.columns)
, '\n--------No. of numerical features:' , len(numerical_cols)
, '\n--------No. of categorical features:' , len(categorical_cols)
, '\n\nTotal no. of evolutionary features:' , len(X_evolFN)
, '\n\nTotal no. of stability features:' , len(X_stability_FN)
, '\n--------Common stabilty cols:' , len(X_common_stability_Fnum)
, '\n--------Foldx cols:' , len(X_foldX_Fnum)
, '\n\nTotal no. of affinity features:' , len(X_affinityFN)
, '\n--------Common affinity cols:' , len(common_affinity_Fnum)
, '\n--------Gene specific affinity cols:' , len(gene_affinity_colnames)
, '\n\nTotal no. of residue level features:', len(X_resprop_FN)
, '\n--------AA index cols:' , len(X_aaindex_Fnum)
, '\n--------Residue Prop cols:' , len(X_str_Fnum)
, '\n--------AA change Prop cols:' , len(X_aap_Fcat)
, '\n\nTotal no. of genomic features:' , len(X_genomicFN)
, '\n--------MAF+OR cols:' , len(X_gn_mafor_Fnum)
, '\n--------Lineage cols:' , len(X_gn_linegae_Fnum)
, '\n--------Other cols:' , len(X_gn_Fcat)
)
else:
print('\nFAIL: numbers mismatch'
, '\nExpected:',len(X_evolFN) + len(X_stability_FN) + len(X_affinityFN) + len(X_resprop_FN) + len(X_genomicFN)
, '\nGot:', len(X.columns))
sys.exit()
###############################################################################
print('\n-------------------------------------------------------------'
, '\nSuccessfully split data: ALL features'
, '\nactual values: training set'
, '\nSplit:', tts_split
#, '\nimputed values: blind test set'
, '\n\nTotal data size:', len(X) + len(X_bts)
, '\n\nTrain data size:', X.shape
, '\ny_train numbers:', yc1
, '\n\nTest data size:', X_bts.shape
, '\ny_test_numbers:', yc2
, '\n\ny_train ratio:',yc1_ratio
, '\ny_test ratio:', yc2_ratio
, '\n-------------------------------------------------------------'
)
##########################################################################
# Quick check
#(X['ligand_affinity_change']==0).sum() == (X['ligand_distance']>10).sum()
for i in range(len(cols_to_mask)):
ind = i+1
print('\nindex:', i, '\nind:', ind)
print('\nMask count check:'
, (my_df_ml[cols_to_mask[i]]==0).sum() == (my_df_ml['ligand_distance']>10).sum()
)
print('Original Data\n', Counter(y)
, 'Data dim:', X.shape)
###########################################################################
#%%
###########################################################################
# RESAMPLING
###########################################################################
#------------------------------
# Simple Random oversampling
# [Numerical + catgeorical]
#------------------------------
oversample = RandomOverSampler(sampling_strategy='minority')
X_ros, y_ros = oversample.fit_resample(X, y)
print('\nSimple Random OverSampling\n', Counter(y_ros))
print(X_ros.shape)
#------------------------------
# Simple Random Undersampling
# [Numerical + catgeorical]
#------------------------------
undersample = RandomUnderSampler(sampling_strategy='majority')
X_rus, y_rus = undersample.fit_resample(X, y)
print('\nSimple Random UnderSampling\n', Counter(y_rus))
print(X_rus.shape)
#------------------------------
# Simple combine ROS and RUS
# [Numerical + catgeorical]
#------------------------------
oversample = RandomOverSampler(sampling_strategy='minority')
X_ros, y_ros = oversample.fit_resample(X, y)
undersample = RandomUnderSampler(sampling_strategy='majority')
X_rouC, y_rouC = undersample.fit_resample(X_ros, y_ros)
print('\nSimple Combined Over and UnderSampling\n', Counter(y_rouC))
print(X_rouC.shape)
#------------------------------
# SMOTE_NC: oversampling
# [numerical + categorical]
#https://stackoverflow.com/questions/47655813/oversampling-smote-for-binary-and-categorical-data-in-python
#------------------------------
# Determine categorical and numerical features
numerical_ix = X.select_dtypes(include=['int64', 'float64']).columns
numerical_ix
num_featuresL = list(numerical_ix)
numerical_colind = X.columns.get_indexer(list(numerical_ix) )
numerical_colind
categorical_ix = X.select_dtypes(include=['object', 'bool']).columns
categorical_ix
categorical_colind = X.columns.get_indexer(list(categorical_ix))
categorical_colind
k_sm = 5 # 5 is default
sm_nc = SMOTENC(categorical_features=categorical_colind, k_neighbors = k_sm, **rs, **njobs)
X_smnc, y_smnc = sm_nc.fit_resample(X, y)
print('\nSMOTE_NC OverSampling\n', Counter(y_smnc))
print(X_smnc.shape)
globals().update(locals()) # TROLOLOLOLOLOLS
#print("i did a horrible hack :-)")
###############################################################################
#%% SMOTE RESAMPLING for NUMERICAL ONLY*
# #------------------------------
# # SMOTE: Oversampling
# # [Numerical ONLY]
# #------------------------------
# k_sm = 1
# sm = SMOTE(sampling_strategy = 'auto', k_neighbors = k_sm, **rs)
# X_sm, y_sm = sm.fit_resample(X, y)
# print(X_sm.shape)
# print('\nSMOTE OverSampling\n', Counter(y_sm))
# y_sm_df = y_sm.to_frame()
# y_sm_df.value_counts().plot(kind = 'bar')
# #------------------------------
# # SMOTE: Over + Undersampling COMBINED
# # [Numerical ONLY]
# #-----------------------------
# sm_enn = SMOTEENN(enn=EditedNearestNeighbours(sampling_strategy='all', **rs, **njobs ))
# X_enn, y_enn = sm_enn.fit_resample(X, y)
# print(X_enn.shape)
# print('\nSMOTE Over+Under Sampling combined\n', Counter(y_enn))
###############################################################################
# TODO: Find over and undersampling JUST for categorical data
###########################################################################
print('\n#################################################################'
, '\nDim of X for gene:', gene.lower(), '\n', X.shape
, '\n###############################################################')

787
scripts/ml/ml_data_fg.py
View file

@ -1,787 +0,0 @@
#!/usr/bin/env python3
# -*- coding: utf-8 -*-
"""
Created on Sun Mar 6 13:41:54 2022
@author: tanu
"""
def setvars(gene,drug):
#https://stackoverflow.com/questions/51695322/compare-multiple-algorithms-with-sklearn-pipeline
import os, sys
import pandas as pd
import numpy as np
print(np.__version__)
print(pd.__version__)
import pprint as pp
from copy import deepcopy
from collections import Counter
from sklearn.impute import KNNImputer as KNN
from imblearn.over_sampling import RandomOverSampler
from imblearn.under_sampling import RandomUnderSampler
from imblearn.over_sampling import SMOTE
from sklearn.datasets import make_classification
from imblearn.combine import SMOTEENN
from imblearn.combine import SMOTETomek
from imblearn.over_sampling import SMOTENC
from imblearn.under_sampling import EditedNearestNeighbours
from imblearn.under_sampling import RepeatedEditedNearestNeighbours
from sklearn.metrics import make_scorer, confusion_matrix, accuracy_score, balanced_accuracy_score, precision_score, average_precision_score, recall_score
from sklearn.metrics import roc_auc_score, roc_curve, f1_score, matthews_corrcoef, jaccard_score, classification_report
from sklearn.model_selection import train_test_split, cross_validate, cross_val_score
from sklearn.model_selection import StratifiedKFold,RepeatedStratifiedKFold, RepeatedKFold
from sklearn.pipeline import Pipeline, make_pipeline
import argparse
import re
#%% GLOBALS
rs = {'random_state': 42}
njobs = {'n_jobs': 10}
scoring_fn = ({ 'mcc' : make_scorer(matthews_corrcoef)
, 'accuracy' : make_scorer(accuracy_score)
, 'fscore' : make_scorer(f1_score)
, 'precision' : make_scorer(precision_score)
, 'recall' : make_scorer(recall_score)
, 'roc_auc' : make_scorer(roc_auc_score)
, 'jcc' : make_scorer(jaccard_score)
})
skf_cv = StratifiedKFold(n_splits = 10
#, shuffle = False, random_state= None)
, shuffle = True,**rs)
rskf_cv = RepeatedStratifiedKFold(n_splits = 10
, n_repeats = 3
, **rs)
mcc_score_fn = {'mcc': make_scorer(matthews_corrcoef)}
jacc_score_fn = {'jcc': make_scorer(jaccard_score)}
#%% FOR LATER: Combine ED logo data
###########################################################################
rs = {'random_state': 42}
njobs = {'n_jobs': 10}
homedir = os.path.expanduser("~")
geneL_basic = ['pnca']
geneL_na = ['gid']
geneL_na_ppi2 = ['rpob']
geneL_ppi2 = ['alr', 'embb', 'katg']
#num_type = ['int64', 'float64']
num_type = ['int16', 'int32', 'int64', 'float16', 'float32', 'float64']
cat_type = ['object', 'bool']
#==============
# directories
#==============
datadir = homedir + '/git/Data/'
indir = datadir + drug + '/input/'
outdir = datadir + drug + '/output/'
#=======
# input
#=======
#---------
# File 1
#---------
infile_ml1 = outdir + gene.lower() + '_merged_df3.csv'
#infile_ml2 = outdir + gene.lower() + '_merged_df2.csv'
my_features_df = pd.read_csv(infile_ml1, index_col = 0)
my_features_df = my_features_df .reset_index(drop = True)
my_features_df.index
my_features_df.dtypes
mycols = my_features_df.columns
#---------
# File 2
#---------
infile_aaindex = outdir + 'aa_index/' + gene.lower() + '_aa.csv'
aaindex_df = pd.read_csv(infile_aaindex, index_col = 0)
aaindex_df.dtypes
#-----------
# check for non-numerical columns
#-----------
if any(aaindex_df.dtypes==object):
print('\naaindex_df contains non-numerical data')
aaindex_df_object = aaindex_df.select_dtypes(include = cat_type)
print('\nTotal no. of non-numerial columns:', len(aaindex_df_object.columns))
expected_aa_ncols = len(aaindex_df.columns) - len(aaindex_df_object.columns)
#-----------
# Extract numerical data only
#-----------
print('\nSelecting numerical data only')
aaindex_df = aaindex_df.select_dtypes(include = num_type)
#---------------------------
# aaindex: sanity check 1
#---------------------------
if len(aaindex_df.columns) == expected_aa_ncols:
print('\nPASS: successfully selected numerical columns only for aaindex_df')
else:
print('\nFAIL: Numbers mismatch'
, '\nExpected ncols:', expected_aa_ncols
, '\nGot:', len(aaindex_df.columns))
#---------------
# check for NA
#---------------
print('\nNow checking for NA in the remaining aaindex_cols')
c1 = aaindex_df.isna().sum()
c2 = c1.sort_values(ascending=False)
print('\nCounting aaindex_df cols with NA'
, '\nncols with NA:', sum(c2>0), 'columns'
, '\nDropping these...'
, '\nOriginal ncols:', len(aaindex_df.columns)
)
aa_df = aaindex_df.dropna(axis=1)
print('\nRevised df ncols:', len(aa_df.columns))
c3 = aa_df.isna().sum()
c4 = c3.sort_values(ascending=False)
print('\nChecking NA in revised df...')
if sum(c4>0):
sys.exit('\nFAIL: aaindex_df still contains cols with NA, please check and drop these before proceeding...')
else:
print('\nPASS: cols with NA successfully dropped from aaindex_df'
, '\nProceeding with combining aa_df with other features_df')
#---------------------------
# aaindex: sanity check 2
#---------------------------
expected_aa_ncols2 = len(aaindex_df.columns) - sum(c2>0)
if len(aa_df.columns) == expected_aa_ncols2:
print('\nPASS: ncols match'
, '\nExpected ncols:', expected_aa_ncols2
, '\nGot:', len(aa_df.columns))
else:
print('\nFAIL: Numbers mismatch'
, '\nExpected ncols:', expected_aa_ncols2
, '\nGot:', len(aa_df.columns))
# Important: need this to identify aaindex cols
aa_df_cols = aa_df.columns
print('\nTotal no. of columns in clean aa_df:', len(aa_df_cols))
###############################################################################
#%% Combining my_features_df and aaindex_df
#===========================
# Merge my_df + aaindex_df
#===========================
if aa_df.columns[aa_df.columns.isin(my_features_df.columns)] == my_features_df.columns[my_features_df.columns.isin(aa_df.columns)]:
print('\nMerging on column: mutationinformation')
if len(my_features_df) == len(aa_df):
expected_nrows = len(my_features_df)
print('\nProceeding to merge, expected nrows in merged_df:', expected_nrows)
else:
sys.exit('\nNrows mismatch, cannot merge. Please check'
, '\nnrows my_df:', len(my_features_df)
, '\nnrows aa_df:', len(aa_df))
#-----------------
# Reset index: mutationinformation
# Very important for merging
#-----------------
aa_df = aa_df.reset_index()
expected_ncols = len(my_features_df.columns) + len(aa_df.columns) - 1 # for the no. of merging col
#-----------------
# Merge: my_features_df + aa_df
#-----------------
merged_df = pd.merge(my_features_df
, aa_df
, on = 'mutationinformation')
#---------------------------
# aaindex: sanity check 3
#---------------------------
if len(merged_df.columns) == expected_ncols:
print('\nPASS: my_features_df and aa_df successfully combined'
, '\nnrows:', len(merged_df)
, '\nncols:', len(merged_df.columns))
else:
sys.exit('\nFAIL: could not combine my_features_df and aa_df'
, '\nCheck dims and merging cols!')
#--------
# Reassign so downstream code doesn't need to change
#--------
my_df = merged_df.copy()
#%% Data: my_df
# Check if non structural pos have crept in
# IDEALLY remove from source! But for rpoB do it here
# Drop NA where numerical cols have them
if gene.lower() in geneL_na_ppi2:
#D1148 get rid of
na_index = my_df['mutationinformation'].index[my_df['mcsm_na_affinity'].apply(np.isnan)]
my_df = my_df.drop(index=na_index)
# FIXED: complete data for all muts inc L114M, F115L, V123L, V125I, V131M
# if gene.lower() in ['embb']:
# na_index = my_df['mutationinformation'].index[my_df['ligand_distance'].apply(np.isnan)]
# my_df = my_df.drop(index=na_index)
# # Sanity check for non-structural positions
# print('\nChecking for non-structural postions')
# na_index = my_df['mutationinformation'].index[my_df['ligand_distance'].apply(np.isnan)]
# if len(na_index) > 0:
# print('\nNon-structural positions detected for gene:', gene.lower()
# , '\nTotal number of these detected:', len(na_index)
# , '\These are at index:', na_index
# , '\nOriginal nrows:', len(my_df)
# , '\nDropping these...')
# my_df = my_df.drop(index=na_index)
# print('\nRevised nrows:', len(my_df))
# else:
# print('\nNo non-structural positions detected for gene:', gene.lower()
# , '\nnrows:', len(my_df))
###########################################################################
#%% Add lineage calculation columns
#FIXME: Check if this can be imported from config?
total_mtblineage_uc = 8
lineage_colnames = ['lineage_list_all', 'lineage_count_all', 'lineage_count_unique', 'lineage_list_unique', 'lineage_multimode']
#bar = my_df[lineage_colnames]
my_df['lineage_proportion'] = my_df['lineage_count_unique']/my_df['lineage_count_all']
my_df['dist_lineage_proportion'] = my_df['lineage_count_unique']/total_mtblineage_uc
###########################################################################
#%% Active site annotation column
# change from numberic to categorical
if my_df['active_site'].dtype in num_type:
my_df['active_site'] = my_df['active_site'].astype(object)
my_df['active_site'].dtype
#%% AA property change
#--------------------
# Water prop change
#--------------------
my_df['water_change'] = my_df['wt_prop_water'] + str('_to_') + my_df['mut_prop_water']
my_df['water_change'].value_counts()
water_prop_changeD = {
'hydrophobic_to_neutral' : 'change'
, 'hydrophobic_to_hydrophobic' : 'no_change'
, 'neutral_to_neutral' : 'no_change'
, 'neutral_to_hydrophobic' : 'change'
, 'hydrophobic_to_hydrophilic' : 'change'
, 'neutral_to_hydrophilic' : 'change'
, 'hydrophilic_to_neutral' : 'change'
, 'hydrophilic_to_hydrophobic' : 'change'
, 'hydrophilic_to_hydrophilic' : 'no_change'
}
my_df['water_change'] = my_df['water_change'].map(water_prop_changeD)
my_df['water_change'].value_counts()
#--------------------
# Polarity change
#--------------------
my_df['polarity_change'] = my_df['wt_prop_polarity'] + str('_to_') + my_df['mut_prop_polarity']
my_df['polarity_change'].value_counts()
polarity_prop_changeD = {
'non-polar_to_non-polar' : 'no_change'
, 'non-polar_to_neutral' : 'change'
, 'neutral_to_non-polar' : 'change'
, 'neutral_to_neutral' : 'no_change'
, 'non-polar_to_basic' : 'change'
, 'acidic_to_neutral' : 'change'
, 'basic_to_neutral' : 'change'
, 'non-polar_to_acidic' : 'change'
, 'neutral_to_basic' : 'change'
, 'acidic_to_non-polar' : 'change'
, 'basic_to_non-polar' : 'change'
, 'neutral_to_acidic' : 'change'
, 'acidic_to_acidic' : 'no_change'
, 'basic_to_acidic' : 'change'
, 'basic_to_basic' : 'no_change'
, 'acidic_to_basic' : 'change'}
my_df['polarity_change'] = my_df['polarity_change'].map(polarity_prop_changeD)
my_df['polarity_change'].value_counts()
#--------------------
# Electrostatics change
#--------------------
my_df['electrostatics_change'] = my_df['wt_calcprop'] + str('_to_') + my_df['mut_calcprop']
my_df['electrostatics_change'].value_counts()
calc_prop_changeD = {
'non-polar_to_non-polar' : 'no_change'
, 'non-polar_to_polar' : 'change'
, 'polar_to_non-polar' : 'change'
, 'non-polar_to_pos' : 'change'
, 'neg_to_non-polar' : 'change'
, 'non-polar_to_neg' : 'change'
, 'pos_to_polar' : 'change'
, 'pos_to_non-polar' : 'change'
, 'polar_to_polar' : 'no_change'
, 'neg_to_neg' : 'no_change'
, 'polar_to_neg' : 'change'
, 'pos_to_neg' : 'change'
, 'pos_to_pos' : 'no_change'
, 'polar_to_pos' : 'change'
, 'neg_to_polar' : 'change'
, 'neg_to_pos' : 'change'
}
my_df['electrostatics_change'] = my_df['electrostatics_change'].map(calc_prop_changeD)
my_df['electrostatics_change'].value_counts()
#--------------------
# Summary change: Create a combined column summarising these three cols
#--------------------
detect_change = 'change'
check_prop_cols = ['water_change', 'polarity_change', 'electrostatics_change']
#my_df['aa_prop_change'] = (my_df.values == detect_change).any(1).astype(int)
my_df['aa_prop_change'] = (my_df[check_prop_cols].values == detect_change).any(1).astype(int)
my_df['aa_prop_change'].value_counts()
my_df['aa_prop_change'].dtype
my_df['aa_prop_change'] = my_df['aa_prop_change'].map({1:'change'
, 0: 'no_change'})
my_df['aa_prop_change'].value_counts()
my_df['aa_prop_change'].dtype
#%% IMPUTE values for OR [check script for exploration: UQ_or_imputer]
#--------------------
# Impute OR values
#--------------------
#or_cols = ['or_mychisq', 'log10_or_mychisq', 'or_fisher']
sel_cols = ['mutationinformation', 'or_mychisq', 'log10_or_mychisq']
or_cols = ['or_mychisq', 'log10_or_mychisq']
print("count of NULL values before imputation\n")
print(my_df[or_cols].isnull().sum())
my_dfI = pd.DataFrame(index = my_df['mutationinformation'] )
my_dfI = pd.DataFrame(KNN(n_neighbors=3, weights="uniform").fit_transform(my_df[or_cols])
, index = my_df['mutationinformation']
, columns = or_cols )
my_dfI.columns = ['or_rawI', 'logorI']
my_dfI.columns
my_dfI = my_dfI.reset_index(drop = False) # prevents old index from being added as a column
my_dfI.head()
print("count of NULL values AFTER imputation\n")
print(my_dfI.isnull().sum())
#-------------------------------------------
# OR df Merge: with original based on index
#-------------------------------------------
#my_df['index_bm'] = my_df.index
mydf_imputed = pd.merge(my_df
, my_dfI
, on = 'mutationinformation')
#mydf_imputed = mydf_imputed.set_index(['index_bm'])
my_df['log10_or_mychisq'].isna().sum()
mydf_imputed['log10_or_mychisq'].isna().sum()
mydf_imputed['logorI'].isna().sum() # should be 0
len(my_df.columns)
len(mydf_imputed.columns)
#-----------------------------------------
# REASSIGN my_df after imputing OR values
#-----------------------------------------
my_df = mydf_imputed.copy()
if my_df['logorI'].isna().sum() == 0:
print('\nPASS: OR values imputed, data ready for ML')
else:
sys.exit('\nFAIL: something went wrong, Data not ready for ML. Please check upstream!')
#!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
#---------------------------------------
# TODO: try other imputation like MICE
#---------------------------------------
#!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
#%%########################################################################
#==========================
# Data for ML
#==========================
my_df_ml = my_df.copy()
# Build column names to mask for affinity chanhes
if gene.lower() in geneL_basic:
#X_stabilityN = common_cols_stabiltyN
gene_affinity_colnames = []# not needed as its the common ones
cols_to_mask = ['ligand_affinity_change']
if gene.lower() in geneL_ppi2:
gene_affinity_colnames = ['mcsm_ppi2_affinity', 'interface_dist']
#X_stabilityN = common_cols_stabiltyN + geneL_ppi2_st_cols
cols_to_mask = ['ligand_affinity_change', 'mcsm_ppi2_affinity']
if gene.lower() in geneL_na:
gene_affinity_colnames = ['mcsm_na_affinity']
#X_stabilityN = common_cols_stabiltyN + geneL_na_st_cols
cols_to_mask = ['ligand_affinity_change', 'mcsm_na_affinity']
if gene.lower() in geneL_na_ppi2:
gene_affinity_colnames = ['mcsm_na_affinity'] + ['mcsm_ppi2_affinity', 'interface_dist']
#X_stabilityN = common_cols_stabiltyN + geneL_na_ppi2_st_cols
cols_to_mask = ['ligand_affinity_change', 'mcsm_na_affinity', 'mcsm_ppi2_affinity']
#=======================
# Masking columns:
# (mCSM-lig, mCSM-NA, mCSM-ppi2) values for lig_dist >10
#=======================
my_df_ml['mutationinformation'][my_df_ml['ligand_distance']>10].value_counts()
my_df_ml.groupby('mutationinformation')['ligand_distance'].apply(lambda x: (x>10)).value_counts()
my_df_ml.loc[(my_df_ml['ligand_distance'] > 10), cols_to_mask].value_counts()
# mask the mcsm affinity related columns where ligand distance > 10
my_df_ml.loc[(my_df_ml['ligand_distance'] > 10), cols_to_mask] = 0
(my_df_ml['ligand_affinity_change'] == 0).sum()
mask_check = my_df_ml[['mutationinformation', 'ligand_distance'] + cols_to_mask]
#===================================================
# write file for check
mask_check.sort_values(by = ['ligand_distance'], ascending = True, inplace = True)
mask_check.to_csv(outdir + 'ml/' + gene.lower() + '_mask_check.csv')
#===================================================
###############################################################################
#%% Feature groups (FG): Build X for Input ML
############################################################################
#===========================
# FG1: Evolutionary features
#===========================
X_evolFN = ['consurf_score'
, 'snap2_score'
, 'provean_score']
###############################################################################
#========================
# FG2: Stability features
#========================
#--------
# common
#--------
X_common_stability_Fnum = [
'duet_stability_change'
, 'ddg_foldx'
, 'deepddg'
, 'ddg_dynamut2'
, 'contacts']
#--------
# FoldX
#--------
X_foldX_Fnum = [ 'electro_rr', 'electro_mm', 'electro_sm', 'electro_ss'
, 'disulfide_rr', 'disulfide_mm', 'disulfide_sm', 'disulfide_ss'
, 'hbonds_rr', 'hbonds_mm', 'hbonds_sm', 'hbonds_ss'
, 'partcov_rr', 'partcov_mm', 'partcov_sm', 'partcov_ss'
, 'vdwclashes_rr', 'vdwclashes_mm', 'vdwclashes_sm', 'vdwclashes_ss'
, 'volumetric_rr', 'volumetric_mm', 'volumetric_ss']
X_stability_FN = X_common_stability_Fnum + X_foldX_Fnum
###############################################################################
#===================
# FG3: Affinity features
#===================
common_affinity_Fnum = ['ligand_distance'
, 'ligand_affinity_change'
, 'mmcsm_lig']
# if gene.lower() in geneL_basic:
# X_affinityFN = common_affinity_Fnum
# else:
# X_affinityFN = common_affinity_Fnum + gene_affinity_colnames
X_affinityFN = common_affinity_Fnum + gene_affinity_colnames
###############################################################################
#============================
# FG4: Residue level features
#============================
#-----------
# AA index
#-----------
X_aaindex_Fnum = list(aa_df_cols)
print('\nTotal no. of features for aaindex:', len(X_aaindex_Fnum))
#-----------------
# surface area
# depth
# hydrophobicity
#-----------------
X_str_Fnum = ['rsa'
#, 'asa'
, 'kd_values'
, 'rd_values']
#---------------------------
# Other aa properties
# active site indication
#---------------------------
X_aap_Fcat = ['ss_class'
# , 'wt_prop_water'
# , 'mut_prop_water'
# , 'wt_prop_polarity'
# , 'mut_prop_polarity'
# , 'wt_calcprop'
# , 'mut_calcprop'
, 'aa_prop_change'
, 'electrostatics_change'
, 'polarity_change'
, 'water_change'
, 'active_site']
X_resprop_FN = X_aaindex_Fnum + X_str_Fnum + X_aap_Fcat
###############################################################################
#========================
# FG5: Genomic features
#========================
X_gn_mafor_Fnum = ['maf'
, 'logorI'
# , 'or_rawI'
# , 'or_mychisq'
# , 'or_logistic'
# , 'or_fisher'
# , 'pval_fisher'
]
X_gn_linegae_Fnum = ['lineage_proportion'
, 'dist_lineage_proportion'
#, 'lineage' # could be included as a category but it has L2;L4 formatting
, 'lineage_count_all'
, 'lineage_count_unique'
]
X_gn_Fcat = ['drtype_mode_labels' # beware then you can't use it to predict [USED it for uq_v1, not v2]
#, 'gene_name' # will be required for the combined stuff
]
X_genomicFN = X_gn_mafor_Fnum + X_gn_linegae_Fnum + X_gn_Fcat
###############################################################################
#========================
# FG6 collapsed: Structural : Atability + Affinity + ResidueProp
#========================
X_structural_FN = X_stability_FN + X_affinityFN + X_resprop_FN
###############################################################################
#========================
# BUILDING all features
#========================
all_featuresN = X_evolFN + X_structural_FN + X_genomicFN
###############################################################################
#%% Define training and test data
#======================================================
# Training and BLIND test set: actual vs imputed
# dst with actual values : training set
# dst with imputed values : blind test
#======================================================
my_df_ml[drug].isna().sum() #'na' ones are the blind_test set
blind_test_df = my_df_ml[my_df_ml[drug].isna()]
blind_test_df.shape
training_df = my_df_ml[my_df_ml[drug].notna()]
training_df.shape
# Target 1: dst_mode
training_df[drug].value_counts()
training_df['dst_mode'].value_counts()
####################################################################
#=====================================
# ML data: actual vs imputed
#=====================================
#------
# X: Training and Blind test (BTS)
#------
X = training_df[all_featuresN]
X_bts = blind_test_df[all_featuresN]
#------
# y
#------
y = training_df['dst_mode']
y_bts = blind_test_df['dst_mode']
yc1 = Counter(y)
yc1_ratio = yc1[0]/yc1[1]
yc2 = Counter(y_bts)
yc2_ratio = yc2[0]/yc2[1]
###############################################################################
#======================================================
# Determine categorical and numerical features
#======================================================
numerical_cols = X.select_dtypes(include=['int64', 'float64']).columns
numerical_cols
categorical_cols = X.select_dtypes(include=['object', 'bool']).columns
categorical_cols
################################################################################
# IMPORTANT sanity checks
if len(X.columns) == len(X_evolFN) + len(X_stability_FN) + len(X_affinityFN) + len(X_resprop_FN) + len(X_genomicFN):
print('\nPASS: ML data with input features, training and test generated...'
, '\n\nTotal no. of input features:' , len(X.columns)
, '\n--------No. of numerical features:' , len(numerical_cols)
, '\n--------No. of categorical features:' , len(categorical_cols)
, '\n\nTotal no. of evolutionary features:' , len(X_evolFN)
, '\n\nTotal no. of stability features:' , len(X_stability_FN)
, '\n--------Common stabilty cols:' , len(X_common_stability_Fnum)
, '\n--------Foldx cols:' , len(X_foldX_Fnum)
, '\n\nTotal no. of affinity features:' , len(X_affinityFN)
, '\n--------Common affinity cols:' , len(common_affinity_Fnum)
, '\n--------Gene specific affinity cols:' , len(gene_affinity_colnames)
, '\n\nTotal no. of residue level features:', len(X_resprop_FN)
, '\n--------AA index cols:' , len(X_aaindex_Fnum)
, '\n--------Residue Prop cols:' , len(X_str_Fnum)
, '\n--------AA change Prop cols:' , len(X_aap_Fcat)
, '\n\nTotal no. of genomic features:' , len(X_genomicFN)
, '\n--------MAF+OR cols:' , len(X_gn_mafor_Fnum)
, '\n--------Lineage cols:' , len(X_gn_linegae_Fnum)
, '\n--------Other cols:' , len(X_gn_Fcat)
)
else:
print('\nFAIL: numbers mismatch'
, '\nExpected:',len(X_evolFN) + len(X_stability_FN) + len(X_affinityFN) + len(X_resprop_FN) + len(X_genomicFN)
, '\nGot:', len(X.columns))
sys.exit()
###############################################################################
print('\n-------------------------------------------------------------'
, '\nSuccessfully split data: ALL features'
, '\nactual values: training set'
, '\nimputed values: blind test set'
, '\n\nTotal data size:', len(X) + len(X_bts)
, '\n\nTrain data size:', X.shape
, '\ny_train numbers:', yc1
, '\n\nTest data size:', X_bts.shape
, '\ny_test_numbers:', yc2
, '\n\ny_train ratio:',yc1_ratio
, '\ny_test ratio:', yc2_ratio
, '\n-------------------------------------------------------------'
)
##########################################################################
# Quick check
#(X['ligand_affinity_change']==0).sum() == (X['ligand_distance']>10).sum()
for i in range(len(cols_to_mask)):
ind = i+1
print('\nindex:', i, '\nind:', ind)
print('\nMask count check:'
, (my_df_ml[cols_to_mask[i]]==0).sum() == (my_df_ml['ligand_distance']>10).sum()
)
print('Original Data\n', Counter(y)
, 'Data dim:', X.shape)
###########################################################################
#%%
###########################################################################
# RESAMPLING
###########################################################################
#------------------------------
# Simple Random oversampling
# [Numerical + catgeorical]
#------------------------------
oversample = RandomOverSampler(sampling_strategy='minority')
X_ros, y_ros = oversample.fit_resample(X, y)
print('Simple Random OverSampling\n', Counter(y_ros))
print(X_ros.shape)
#------------------------------
# Simple Random Undersampling
# [Numerical + catgeorical]
#------------------------------
undersample = RandomUnderSampler(sampling_strategy='majority')
X_rus, y_rus = undersample.fit_resample(X, y)
print('Simple Random UnderSampling\n', Counter(y_rus))
print(X_rus.shape)
#------------------------------
# Simple combine ROS and RUS
# [Numerical + catgeorical]
#------------------------------
oversample = RandomOverSampler(sampling_strategy='minority')
X_ros, y_ros = oversample.fit_resample(X, y)
undersample = RandomUnderSampler(sampling_strategy='majority')
X_rouC, y_rouC = undersample.fit_resample(X_ros, y_ros)
print('Simple Combined Over and UnderSampling\n', Counter(y_rouC))
print(X_rouC.shape)
#------------------------------
# SMOTE_NC: oversampling
# [numerical + categorical]
#https://stackoverflow.com/questions/47655813/oversampling-smote-for-binary-and-categorical-data-in-python
#------------------------------
# Determine categorical and numerical features
numerical_ix = X.select_dtypes(include=['int64', 'float64']).columns
numerical_ix
num_featuresL = list(numerical_ix)
numerical_colind = X.columns.get_indexer(list(numerical_ix) )
numerical_colind
categorical_ix = X.select_dtypes(include=['object', 'bool']).columns
categorical_ix
categorical_colind = X.columns.get_indexer(list(categorical_ix))
categorical_colind
k_sm = 5 # 5 is deafult
sm_nc = SMOTENC(categorical_features=categorical_colind, k_neighbors = k_sm, **rs, **njobs)
X_smnc, y_smnc = sm_nc.fit_resample(X, y)
print('\nSMOTE_NC OverSampling\n', Counter(y_smnc))
print(X_smnc.shape)
globals().update(locals()) # TROLOLOLOLOLOLS
#print("i did a horrible hack :-)")
###############################################################################
#%% SMOTE RESAMPLING for NUMERICAL ONLY*
# #------------------------------
# # SMOTE: Oversampling
# # [Numerical ONLY]
# #------------------------------
# k_sm = 1
# sm = SMOTE(sampling_strategy = 'auto', k_neighbors = k_sm, **rs)
# X_sm, y_sm = sm.fit_resample(X, y)
# print(X_sm.shape)
# print('\nSMOTE OverSampling\n', Counter(y_sm))
# y_sm_df = y_sm.to_frame()
# y_sm_df.value_counts().plot(kind = 'bar')
# #------------------------------
# # SMOTE: Over + Undersampling COMBINED
# # [Numerical ONLY]
# #-----------------------------
# sm_enn = SMOTEENN(enn=EditedNearestNeighbours(sampling_strategy='all', **rs, **njobs ))
# X_enn, y_enn = sm_enn.fit_resample(X, y)
# print(X_enn.shape)
# print('\nSMOTE Over+Under Sampling combined\n', Counter(y_enn))
###############################################################################
# TODO: Find over and undersampling JUST for categorical data

700
scripts/ml/ml_data_orig.py
View file

@ -1,700 +0,0 @@
#!/usr/bin/env python3
# -*- coding: utf-8 -*-
"""
Created on Sun Mar 6 13:41:54 2022
@author: tanu
"""
def setvars(gene,drug):
#https://stackoverflow.com/questions/51695322/compare-multiple-algorithms-with-sklearn-pipeline
import os, sys
import pandas as pd
import numpy as np
print(np.__version__)
print(pd.__version__)
import pprint as pp
from copy import deepcopy
from collections import Counter
from sklearn.impute import KNNImputer as KNN
from imblearn.over_sampling import RandomOverSampler
from imblearn.under_sampling import RandomUnderSampler
from imblearn.over_sampling import SMOTE
from sklearn.datasets import make_classification
from imblearn.combine import SMOTEENN
from imblearn.combine import SMOTETomek
from imblearn.over_sampling import SMOTENC
from imblearn.under_sampling import EditedNearestNeighbours
from imblearn.under_sampling import RepeatedEditedNearestNeighbours
from sklearn.metrics import make_scorer, confusion_matrix, accuracy_score, balanced_accuracy_score, precision_score, average_precision_score, recall_score
from sklearn.metrics import roc_auc_score, roc_curve, f1_score, matthews_corrcoef, jaccard_score, classification_report
from sklearn.model_selection import train_test_split, cross_validate, cross_val_score
from sklearn.model_selection import StratifiedKFold,RepeatedStratifiedKFold, RepeatedKFold
from sklearn.pipeline import Pipeline, make_pipeline
#%% GLOBALS
rs = {'random_state': 42}
njobs = {'n_jobs': 10}
scoring_fn = ({ 'mcc' : make_scorer(matthews_corrcoef)
, 'accuracy' : make_scorer(accuracy_score)
, 'fscore' : make_scorer(f1_score)
, 'precision' : make_scorer(precision_score)
, 'recall' : make_scorer(recall_score)
, 'roc_auc' : make_scorer(roc_auc_score)
, 'jcc' : make_scorer(jaccard_score)
})
skf_cv = StratifiedKFold(n_splits = 10
#, shuffle = False, random_state= None)
, shuffle = True,**rs)
rskf_cv = RepeatedStratifiedKFold(n_splits = 10
, n_repeats = 3
, **rs)
mcc_score_fn = {'mcc': make_scorer(matthews_corrcoef)}
jacc_score_fn = {'jcc': make_scorer(jaccard_score)}
#%% FOR LATER: Combine ED logo data
#%% DONE: active aa site annotations **DONE on 15/05/2022 as part of generating merged_dfs
###########################################################################
rs = {'random_state': 42}
njobs = {'n_jobs': 10}
homedir = os.path.expanduser("~")
geneL_basic = ['pnca']
geneL_na = ['gid']
geneL_na_ppi2 = ['rpob']
geneL_ppi2 = ['alr', 'embb', 'katg']
#num_type = ['int64', 'float64']
num_type = ['int16', 'int32', 'int64', 'float16', 'float32', 'float64']
cat_type = ['object', 'bool']
#==============
# directories
#==============
datadir = homedir + '/git/Data/'
indir = datadir + drug + '/input/'
outdir = datadir + drug + '/output/'
#=======
# input
#=======
#---------
# File 1
#---------
infile_ml1 = outdir + gene.lower() + '_merged_df3.csv'
#infile_ml2 = outdir + gene.lower() + '_merged_df2.csv'
my_features_df = pd.read_csv(infile_ml1, index_col = 0)
my_features_df = my_features_df .reset_index(drop = True)
my_features_df.index
my_features_df.dtypes
mycols = my_features_df.columns
#---------
# File 2
#---------
infile_aaindex = outdir + 'aa_index/' + gene.lower() + '_aa.csv'
aaindex_df = pd.read_csv(infile_aaindex, index_col = 0)
aaindex_df.dtypes
#-----------
# check for non-numerical columns
#-----------
if any(aaindex_df.dtypes==object):
print('\naaindex_df contains non-numerical data')
aaindex_df_object = aaindex_df.select_dtypes(include = cat_type)
print('\nTotal no. of non-numerial columns:', len(aaindex_df_object.columns))
expected_aa_ncols = len(aaindex_df.columns) - len(aaindex_df_object.columns)
#-----------
# Extract numerical data only
#-----------
print('\nSelecting numerical data only')
aaindex_df = aaindex_df.select_dtypes(include = num_type)
#---------------------------
# aaindex: sanity check 1
#---------------------------
if len(aaindex_df.columns) == expected_aa_ncols:
print('\nPASS: successfully selected numerical columns only for aaindex_df')
else:
print('\nFAIL: Numbers mismatch'
, '\nExpected ncols:', expected_aa_ncols
, '\nGot:', len(aaindex_df.columns))
#---------------
# check for NA
#---------------
print('\nNow checking for NA in the remaining aaindex_cols')
c1 = aaindex_df.isna().sum()
c2 = c1.sort_values(ascending=False)
print('\nCounting aaindex_df cols with NA'
, '\nncols with NA:', sum(c2>0), 'columns'
, '\nDropping these...'
, '\nOriginal ncols:', len(aaindex_df.columns)
)
aa_df = aaindex_df.dropna(axis=1)
print('\nRevised df ncols:', len(aa_df.columns))
c3 = aa_df.isna().sum()
c4 = c3.sort_values(ascending=False)
print('\nChecking NA in revised df...')
if sum(c4>0):
sys.exit('\nFAIL: aaindex_df still contains cols with NA, please check and drop these before proceeding...')
else:
print('\nPASS: cols with NA successfully dropped from aaindex_df'
, '\nProceeding with combining aa_df with other features_df')
#---------------------------
# aaindex: sanity check 2
#---------------------------
expected_aa_ncols2 = len(aaindex_df.columns) - sum(c2>0)
if len(aa_df.columns) == expected_aa_ncols2:
print('\nPASS: ncols match'
, '\nExpected ncols:', expected_aa_ncols2
, '\nGot:', len(aa_df.columns))
else:
print('\nFAIL: Numbers mismatch'
, '\nExpected ncols:', expected_aa_ncols2
, '\nGot:', len(aa_df.columns))
# Important: need this to identify aaindex cols
aa_df_cols = aa_df.columns
print('\nTotal no. of columns in clean aa_df:', len(aa_df_cols))
###############################################################################
#%% Combining my_features_df and aaindex_df
#===========================
# Merge my_df + aaindex_df
#===========================
if aa_df.columns[aa_df.columns.isin(my_features_df.columns)] == my_features_df.columns[my_features_df.columns.isin(aa_df.columns)]:
print('\nMerging on column: mutationinformation')
if len(my_features_df) == len(aa_df):
expected_nrows = len(my_features_df)
print('\nProceeding to merge, expected nrows in merged_df:', expected_nrows)
else:
sys.exit('\nNrows mismatch, cannot merge. Please check'
, '\nnrows my_df:', len(my_features_df)
, '\nnrows aa_df:', len(aa_df))
#-----------------
# Reset index: mutationinformation
# Very important for merging
#-----------------
aa_df = aa_df.reset_index()
expected_ncols = len(my_features_df.columns) + len(aa_df.columns) - 1 # for the no. of merging col
#-----------------
# Merge: my_features_df + aa_df
#-----------------
merged_df = pd.merge(my_features_df
, aa_df
, on = 'mutationinformation')
#---------------------------
# aaindex: sanity check 3
#---------------------------
if len(merged_df.columns) == expected_ncols:
print('\nPASS: my_features_df and aa_df successfully combined'
, '\nnrows:', len(merged_df)
, '\nncols:', len(merged_df.columns))
else:
sys.exit('\nFAIL: could not combine my_features_df and aa_df'
, '\nCheck dims and merging cols!')
#--------
# Reassign so downstream code doesn't need to change
#--------
my_df = merged_df.copy()
#%% Data: my_df
# Check if non structural pos have crept in
# IDEALLY remove from source! But for rpoB do it here
# Drop NA where numerical cols have them
if gene.lower() in geneL_na_ppi2:
#D1148 get rid of
na_index = my_df['mutationinformation'].index[my_df['mcsm_na_affinity'].apply(np.isnan)]
my_df = my_df.drop(index=na_index)
# FIXED: complete data for all muts inc L114M, F115L, V123L, V125I, V131M
# if gene.lower() in ['embb']:
# na_index = my_df['mutationinformation'].index[my_df['ligand_distance'].apply(np.isnan)]
# my_df = my_df.drop(index=na_index)
# # Sanity check for non-structural positions
# print('\nChecking for non-structural postions')
# na_index = my_df['mutationinformation'].index[my_df['ligand_distance'].apply(np.isnan)]
# if len(na_index) > 0:
# print('\nNon-structural positions detected for gene:', gene.lower()
# , '\nTotal number of these detected:', len(na_index)
# , '\These are at index:', na_index
# , '\nOriginal nrows:', len(my_df)
# , '\nDropping these...')
# my_df = my_df.drop(index=na_index)
# print('\nRevised nrows:', len(my_df))
# else:
# print('\nNo non-structural positions detected for gene:', gene.lower()
# , '\nnrows:', len(my_df))
###########################################################################
#%% Add lineage calculation columns
#FIXME: Check if this can be imported from config?
total_mtblineage_uc = 8
lineage_colnames = ['lineage_list_all', 'lineage_count_all', 'lineage_count_unique', 'lineage_list_unique', 'lineage_multimode']
#bar = my_df[lineage_colnames]
my_df['lineage_proportion'] = my_df['lineage_count_unique']/my_df['lineage_count_all']
my_df['dist_lineage_proportion'] = my_df['lineage_count_unique']/total_mtblineage_uc
###########################################################################
#%% Active site annotation column
# change from numberic to categorical
if my_df['active_site'].dtype in num_type:
my_df['active_site'] = my_df['active_site'].astype(object)
my_df['active_site'].dtype
#%% AA property change
#--------------------
# Water prop change
#--------------------
my_df['water_change'] = my_df['wt_prop_water'] + str('_to_') + my_df['mut_prop_water']
my_df['water_change'].value_counts()
water_prop_changeD = {
'hydrophobic_to_neutral' : 'change'
, 'hydrophobic_to_hydrophobic' : 'no_change'
, 'neutral_to_neutral' : 'no_change'
, 'neutral_to_hydrophobic' : 'change'
, 'hydrophobic_to_hydrophilic' : 'change'
, 'neutral_to_hydrophilic' : 'change'
, 'hydrophilic_to_neutral' : 'change'
, 'hydrophilic_to_hydrophobic' : 'change'
, 'hydrophilic_to_hydrophilic' : 'no_change'
}
my_df['water_change'] = my_df['water_change'].map(water_prop_changeD)
my_df['water_change'].value_counts()
#--------------------
# Polarity change
#--------------------
my_df['polarity_change'] = my_df['wt_prop_polarity'] + str('_to_') + my_df['mut_prop_polarity']
my_df['polarity_change'].value_counts()
polarity_prop_changeD = {
'non-polar_to_non-polar' : 'no_change'
, 'non-polar_to_neutral' : 'change'
, 'neutral_to_non-polar' : 'change'
, 'neutral_to_neutral' : 'no_change'
, 'non-polar_to_basic' : 'change'
, 'acidic_to_neutral' : 'change'
, 'basic_to_neutral' : 'change'
, 'non-polar_to_acidic' : 'change'
, 'neutral_to_basic' : 'change'
, 'acidic_to_non-polar' : 'change'
, 'basic_to_non-polar' : 'change'
, 'neutral_to_acidic' : 'change'
, 'acidic_to_acidic' : 'no_change'
, 'basic_to_acidic' : 'change'
, 'basic_to_basic' : 'no_change'
, 'acidic_to_basic' : 'change'}
my_df['polarity_change'] = my_df['polarity_change'].map(polarity_prop_changeD)
my_df['polarity_change'].value_counts()
#--------------------
# Electrostatics change
#--------------------
my_df['electrostatics_change'] = my_df['wt_calcprop'] + str('_to_') + my_df['mut_calcprop']
my_df['electrostatics_change'].value_counts()
calc_prop_changeD = {
'non-polar_to_non-polar' : 'no_change'
, 'non-polar_to_polar' : 'change'
, 'polar_to_non-polar' : 'change'
, 'non-polar_to_pos' : 'change'
, 'neg_to_non-polar' : 'change'
, 'non-polar_to_neg' : 'change'
, 'pos_to_polar' : 'change'
, 'pos_to_non-polar' : 'change'
, 'polar_to_polar' : 'no_change'
, 'neg_to_neg' : 'no_change'
, 'polar_to_neg' : 'change'
, 'pos_to_neg' : 'change'
, 'pos_to_pos' : 'no_change'
, 'polar_to_pos' : 'change'
, 'neg_to_polar' : 'change'
, 'neg_to_pos' : 'change'
}
my_df['electrostatics_change'] = my_df['electrostatics_change'].map(calc_prop_changeD)
my_df['electrostatics_change'].value_counts()
#--------------------
# Summary change: Create a combined column summarising these three cols
#--------------------
detect_change = 'change'
check_prop_cols = ['water_change', 'polarity_change', 'electrostatics_change']
#my_df['aa_prop_change'] = (my_df.values == detect_change).any(1).astype(int)
my_df['aa_prop_change'] = (my_df[check_prop_cols].values == detect_change).any(1).astype(int)
my_df['aa_prop_change'].value_counts()
my_df['aa_prop_change'].dtype
my_df['aa_prop_change'] = my_df['aa_prop_change'].map({1:'change'
, 0: 'no_change'})
my_df['aa_prop_change'].value_counts()
my_df['aa_prop_change'].dtype
#%% IMPUTE values for OR [check script for exploration: UQ_or_imputer]
#--------------------
# Impute OR values
#--------------------
#or_cols = ['or_mychisq', 'log10_or_mychisq', 'or_fisher']
sel_cols = ['mutationinformation', 'or_mychisq', 'log10_or_mychisq']
or_cols = ['or_mychisq', 'log10_or_mychisq']
print("count of NULL values before imputation\n")
print(my_df[or_cols].isnull().sum())
my_dfI = pd.DataFrame(index = my_df['mutationinformation'] )
my_dfI = pd.DataFrame(KNN(n_neighbors=3, weights="uniform").fit_transform(my_df[or_cols])
, index = my_df['mutationinformation']
, columns = or_cols )
my_dfI.columns = ['or_rawI', 'logorI']
my_dfI.columns
my_dfI = my_dfI.reset_index(drop = False) # prevents old index from being added as a column
my_dfI.head()
print("count of NULL values AFTER imputation\n")
print(my_dfI.isnull().sum())
#-------------------------------------------
# OR df Merge: with original based on index
#-------------------------------------------
#my_df['index_bm'] = my_df.index
mydf_imputed = pd.merge(my_df
, my_dfI
, on = 'mutationinformation')
#mydf_imputed = mydf_imputed.set_index(['index_bm'])
my_df['log10_or_mychisq'].isna().sum()
mydf_imputed['log10_or_mychisq'].isna().sum()
mydf_imputed['logorI'].isna().sum() # should be 0
len(my_df.columns)
len(mydf_imputed.columns)
#-----------------------------------------
# REASSIGN my_df after imputing OR values
#-----------------------------------------
my_df = mydf_imputed.copy()
if my_df['logorI'].isna().sum() == 0:
print('\nPASS: OR values imputed, data ready for ML')
else:
sys.exit('\nFAIL: something went wrong, Data not ready for ML. Please check upstream!')
#!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
#---------------------------------------
# TODO: try other imputation like MICE
#---------------------------------------
#!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
#%%########################################################################
#==========================
# Data for ML
#==========================
my_df_ml = my_df.copy()
#%% Build X: input for ML
common_cols_stabiltyN = ['ligand_distance'
, 'ligand_affinity_change'
, 'duet_stability_change'
, 'ddg_foldx'
, 'deepddg'
, 'ddg_dynamut2'
, 'mmcsm_lig'
, 'contacts']
# Build stability columns ~ gene
if gene.lower() in geneL_basic:
X_stabilityN = common_cols_stabiltyN
cols_to_mask = ['ligand_affinity_change']
if gene.lower() in geneL_ppi2:
# X_stabilityN = common_cols_stabiltyN + ['mcsm_ppi2_affinity' , 'interface_dist']
geneL_ppi2_st_cols = ['mcsm_ppi2_affinity', 'interface_dist']
X_stabilityN = common_cols_stabiltyN + geneL_ppi2_st_cols
cols_to_mask = ['ligand_affinity_change', 'mcsm_ppi2_affinity']
if gene.lower() in geneL_na:
# X_stabilityN = common_cols_stabiltyN + ['mcsm_na_affinity']
geneL_na_st_cols = ['mcsm_na_affinity']
X_stabilityN = common_cols_stabiltyN + geneL_na_st_cols
cols_to_mask = ['ligand_affinity_change', 'mcsm_na_affinity']
if gene.lower() in geneL_na_ppi2:
# X_stabilityN = common_cols_stabiltyN + ['mcsm_na_affinity'] + ['mcsm_ppi2_affinity', 'interface_dist']
geneL_na_ppi2_st_cols = ['mcsm_na_affinity'] + ['mcsm_ppi2_affinity', 'interface_dist']
X_stabilityN = common_cols_stabiltyN + geneL_na_ppi2_st_cols
cols_to_mask = ['ligand_affinity_change', 'mcsm_na_affinity', 'mcsm_ppi2_affinity']
X_foldX_cols = [ 'electro_rr', 'electro_mm', 'electro_sm', 'electro_ss'
, 'disulfide_rr', 'disulfide_mm', 'disulfide_sm', 'disulfide_ss'
, 'hbonds_rr', 'hbonds_mm', 'hbonds_sm', 'hbonds_ss'
, 'partcov_rr', 'partcov_mm', 'partcov_sm', 'partcov_ss'
, 'vdwclashes_rr', 'vdwclashes_mm', 'vdwclashes_sm', 'vdwclashes_ss'
, 'volumetric_rr', 'volumetric_mm', 'volumetric_ss'
]
X_str = ['rsa'
#, 'asa'
, 'kd_values'
, 'rd_values']
X_ssFN = X_stabilityN + X_str + X_foldX_cols
X_evolFN = ['consurf_score'
, 'snap2_score'
, 'provean_score']
X_genomic_mafor = ['maf'
, 'logorI'
# , 'or_rawI'
# , 'or_mychisq'
# , 'or_logistic'
# , 'or_fisher'
# , 'pval_fisher'
]
X_genomic_linegae = ['lineage_proportion'
, 'dist_lineage_proportion'
#, 'lineage' # could be included as a category but it has L2;L4 formatting
, 'lineage_count_all'
, 'lineage_count_unique'
]
X_genomicFN = X_genomic_mafor + X_genomic_linegae
X_aaindexFN = list(aa_df_cols)
print('\nTotal no. of features for aaindex:', len(X_aaindexFN))
# numerical feature names
numerical_FN = X_ssFN + X_evolFN + X_genomicFN + X_aaindexFN
# categorical feature names
categorical_FN = ['ss_class'
# , 'wt_prop_water'
# , 'mut_prop_water'
# , 'wt_prop_polarity'
# , 'mut_prop_polarity'
# , 'wt_calcprop'
# , 'mut_calcprop'
, 'aa_prop_change'
, 'electrostatics_change'
, 'polarity_change'
, 'water_change'
, 'drtype_mode_labels' # beware then you can't use it to predict [USED it for uq_v1, not v2]
, 'active_site' #[didn't use it for uq_v1]
#, 'gene_name' # will be required for the combined stuff
]
#----------------------------------------------
# count numerical and categorical features
#----------------------------------------------
print('\nNo. of numerical features:', len(numerical_FN)
, '\nNo. of categorical features:', len(categorical_FN))
###########################################################################
#=======================
# Masking columns:
# (mCSM-lig, mCSM-NA, mCSM-ppi2) values for lig_dist >10
#=======================
# my_df_ml['mutationinformation'][my_df['ligand_distance']>10].value_counts()
# my_df_ml.groupby('mutationinformation')['ligand_distance'].apply(lambda x: (x>10)).value_counts()
# my_df_ml.loc[(my_df_ml['ligand_distance'] > 10), 'ligand_affinity_change'] = 0
# (my_df_ml['ligand_affinity_change'] == 0).sum()
my_df_ml['mutationinformation'][my_df_ml['ligand_distance']>10].value_counts()
my_df_ml.groupby('mutationinformation')['ligand_distance'].apply(lambda x: (x>10)).value_counts()
my_df_ml.loc[(my_df_ml['ligand_distance'] > 10), cols_to_mask].value_counts()
# mask the mcsm affinity related columns where ligand distance > 10
my_df_ml.loc[(my_df_ml['ligand_distance'] > 10), cols_to_mask] = 0
(my_df_ml['ligand_affinity_change'] == 0).sum()
mask_check = my_df_ml[['mutationinformation', 'ligand_distance'] + cols_to_mask]
# write file for check
mask_check.sort_values(by = ['ligand_distance'], ascending = True, inplace = True)
mask_check.to_csv(outdir + 'ml/' + gene.lower() + '_mask_check.csv')
#===================================================
# Training and BLIND test set: actual vs imputed
# ORIGINAL i.e.
# dst with actual values : training set
# dst with imputed values : blind test
#==================================================
my_df_ml[drug].isna().sum() #'na' ones are the blind_test set
blind_test_df = my_df_ml[my_df_ml[drug].isna()]
blind_test_df.shape
training_df = my_df_ml[my_df_ml[drug].notna()]
training_df.shape
# Target 1: dst_mode
training_df[drug].value_counts()
training_df['dst_mode'].value_counts()
####################################################################
#============
# ML data
#============
#------
# X: Training and Blind test (BTS)
#------
X = training_df[numerical_FN + categorical_FN] # training data ALL
X_bts = blind_test_df[numerical_FN + categorical_FN] # blind test data ALL
#------
# y
#------
y = training_df['dst_mode'] # training data y
y_bts = blind_test_df['dst_mode'] # blind data test y
# Quick check
#(X['ligand_affinity_change']==0).sum() == (X['ligand_distance']>10).sum()
for i in range(len(cols_to_mask)):
ind = i+1
print('\nindex:', i, '\nind:', ind)
print('\nMask count check:'
, (my_df_ml[cols_to_mask[i]]==0).sum() == (my_df_ml['ligand_distance']>10).sum()
)
print('Original Data\n', Counter(y)
, 'Data dim:', X.shape)
yc1 = Counter(y)
yc1_ratio = yc1[0]/yc1[1]
yc2 = Counter(y_bts)
yc2_ratio = yc2[0]/yc2[1]
print('\n-------------------------------------------------------------'
, '\nSuccessfully split data: ORIGINAL training'
, '\nactual values: training set'
, '\nimputed values: blind test set'
, '\nTrain data size:', X.shape
, '\nTest data size:', X_bts.shape
, '\ny_train numbers:', yc1
, '\ny_train ratio:',yc1_ratio
, '\n'
, '\ny_test_numbers:', yc2
, '\ny_test ratio:', yc2_ratio
, '\n-------------------------------------------------------------'
)
###########################################################################
#%%
###########################################################################
# RESAMPLING
###########################################################################
#------------------------------
# Simple Random oversampling
# [Numerical + catgeorical]
#------------------------------
oversample = RandomOverSampler(sampling_strategy='minority')
X_ros, y_ros = oversample.fit_resample(X, y)
print('Simple Random OverSampling\n', Counter(y_ros))
print(X_ros.shape)
#------------------------------
# Simple Random Undersampling
# [Numerical + catgeorical]
#------------------------------
undersample = RandomUnderSampler(sampling_strategy='majority')
X_rus, y_rus = undersample.fit_resample(X, y)
print('Simple Random UnderSampling\n', Counter(y_rus))
print(X_rus.shape)
#------------------------------
# Simple combine ROS and RUS
# [Numerical + catgeorical]
#------------------------------
oversample = RandomOverSampler(sampling_strategy='minority')
X_ros, y_ros = oversample.fit_resample(X, y)
undersample = RandomUnderSampler(sampling_strategy='majority')
X_rouC, y_rouC = undersample.fit_resample(X_ros, y_ros)
print('Simple Combined Over and UnderSampling\n', Counter(y_rouC))
print(X_rouC.shape)
#------------------------------
# SMOTE_NC: oversampling
# [numerical + categorical]
#https://stackoverflow.com/questions/47655813/oversampling-smote-for-binary-and-categorical-data-in-python
#------------------------------
# Determine categorical and numerical features
numerical_ix = X.select_dtypes(include=['int64', 'float64']).columns
numerical_ix
num_featuresL = list(numerical_ix)
numerical_colind = X.columns.get_indexer(list(numerical_ix) )
numerical_colind
categorical_ix = X.select_dtypes(include=['object', 'bool']).columns
categorical_ix
categorical_colind = X.columns.get_indexer(list(categorical_ix))
categorical_colind
k_sm = 5 # 5 is deafult
sm_nc = SMOTENC(categorical_features=categorical_colind, k_neighbors = k_sm, **rs, **njobs)
X_smnc, y_smnc = sm_nc.fit_resample(X, y)
print('SMOTE_NC OverSampling\n', Counter(y_smnc))
print(X_smnc.shape)
globals().update(locals()) # TROLOLOLOLOLOLS
#print("i did a horrible hack :-)")
###############################################################################
#%% SMOTE RESAMPLING for NUMERICAL ONLY*
# #------------------------------
# # SMOTE: Oversampling
# # [Numerical ONLY]
# #------------------------------
# k_sm = 1
# sm = SMOTE(sampling_strategy = 'auto', k_neighbors = k_sm, **rs)
# X_sm, y_sm = sm.fit_resample(X, y)
# print(X_sm.shape)
# print('SMOTE OverSampling\n', Counter(y_sm))
# y_sm_df = y_sm.to_frame()
# y_sm_df.value_counts().plot(kind = 'bar')
# #------------------------------
# # SMOTE: Over + Undersampling COMBINED
# # [Numerical ONLY]
# #-----------------------------
# sm_enn = SMOTEENN(enn=EditedNearestNeighbours(sampling_strategy='all', **rs, **njobs ))
# X_enn, y_enn = sm_enn.fit_resample(X, y)
# print(X_enn.shape)
# print('SMOTE Over+Under Sampling combined\n', Counter(y_enn))
###############################################################################
# TODO: Find over and undersampling JUST for categorical data

735
scripts/ml/ml_data_rt.py
View file

@ -1,735 +0,0 @@
#!/usr/bin/env python3
# -*- coding: utf-8 -*-
"""
Created on Sun Mar 6 13:41:54 2022
@author: tanu
"""
def setvars(gene,drug):
#https://stackoverflow.com/questions/51695322/compare-multiple-algorithms-with-sklearn-pipeline
import os, sys
import pandas as pd
import numpy as np
print(np.__version__)
print(pd.__version__)
import pprint as pp
from copy import deepcopy
from collections import Counter
from sklearn.impute import KNNImputer as KNN
from imblearn.over_sampling import RandomOverSampler
from imblearn.under_sampling import RandomUnderSampler
from imblearn.over_sampling import SMOTE
from sklearn.datasets import make_classification
from imblearn.combine import SMOTEENN
from imblearn.combine import SMOTETomek
from imblearn.over_sampling import SMOTENC
from imblearn.under_sampling import EditedNearestNeighbours
from imblearn.under_sampling import RepeatedEditedNearestNeighbours
from sklearn.metrics import make_scorer, confusion_matrix, accuracy_score, balanced_accuracy_score, precision_score, average_precision_score, recall_score
from sklearn.metrics import roc_auc_score, roc_curve, f1_score, matthews_corrcoef, jaccard_score, classification_report
from sklearn.model_selection import train_test_split, cross_validate, cross_val_score
from sklearn.model_selection import StratifiedKFold,RepeatedStratifiedKFold, RepeatedKFold
from sklearn.pipeline import Pipeline, make_pipeline
#%% GLOBALS
rs = {'random_state': 42}
njobs = {'n_jobs': 10}
scoring_fn = ({ 'mcc' : make_scorer(matthews_corrcoef)
, 'accuracy' : make_scorer(accuracy_score)
, 'fscore' : make_scorer(f1_score)
, 'precision' : make_scorer(precision_score)
, 'recall' : make_scorer(recall_score)
, 'roc_auc' : make_scorer(roc_auc_score)
, 'jcc' : make_scorer(jaccard_score)
})
skf_cv = StratifiedKFold(n_splits = 10
#, shuffle = False, random_state= None)
, shuffle = True,**rs)
rskf_cv = RepeatedStratifiedKFold(n_splits = 10
, n_repeats = 3
, **rs)
mcc_score_fn = {'mcc': make_scorer(matthews_corrcoef)}
jacc_score_fn = {'jcc': make_scorer(jaccard_score)}
#%% FOR LATER: Combine ED logo data
#%% DONE: active aa site annotations **DONE on 15/05/2022 as part of generating merged_dfs
###########################################################################
rs = {'random_state': 42}
njobs = {'n_jobs': 10}
homedir = os.path.expanduser("~")
geneL_basic = ['pnca']
geneL_na = ['gid']
geneL_na_ppi2 = ['rpob']
geneL_ppi2 = ['alr', 'embb', 'katg']
#num_type = ['int64', 'float64']
num_type = ['int16', 'int32', 'int64', 'float16', 'float32', 'float64']
cat_type = ['object', 'bool']
#==============
# directories
#==============
datadir = homedir + '/git/Data/'
indir = datadir + drug + '/input/'
outdir = datadir + drug + '/output/'
#=======
# input
#=======
#---------
# File 1
#---------
infile_ml1 = outdir + gene.lower() + '_merged_df3.csv'
#infile_ml2 = outdir + gene.lower() + '_merged_df2.csv'
my_features_df = pd.read_csv(infile_ml1, index_col = 0)
my_features_df = my_features_df .reset_index(drop = True)
my_features_df.index
my_features_df.dtypes
mycols = my_features_df.columns
#---------
# File 2
#---------
infile_aaindex = outdir + 'aa_index/' + gene.lower() + '_aa.csv'
aaindex_df = pd.read_csv(infile_aaindex, index_col = 0)
aaindex_df.dtypes
#-----------
# check for non-numerical columns
#-----------
if any(aaindex_df.dtypes==object):
print('\naaindex_df contains non-numerical data')
aaindex_df_object = aaindex_df.select_dtypes(include = cat_type)
print('\nTotal no. of non-numerial columns:', len(aaindex_df_object.columns))
expected_aa_ncols = len(aaindex_df.columns) - len(aaindex_df_object.columns)
#-----------
# Extract numerical data only
#-----------
print('\nSelecting numerical data only')
aaindex_df = aaindex_df.select_dtypes(include = num_type)
#---------------------------
# aaindex: sanity check 1
#---------------------------
if len(aaindex_df.columns) == expected_aa_ncols:
print('\nPASS: successfully selected numerical columns only for aaindex_df')
else:
print('\nFAIL: Numbers mismatch'
, '\nExpected ncols:', expected_aa_ncols
, '\nGot:', len(aaindex_df.columns))
#---------------
# check for NA
#---------------
print('\nNow checking for NA in the remaining aaindex_cols')
c1 = aaindex_df.isna().sum()
c2 = c1.sort_values(ascending=False)
print('\nCounting aaindex_df cols with NA'
, '\nncols with NA:', sum(c2>0), 'columns'
, '\nDropping these...'
, '\nOriginal ncols:', len(aaindex_df.columns)
)
aa_df = aaindex_df.dropna(axis=1)
print('\nRevised df ncols:', len(aa_df.columns))
c3 = aa_df.isna().sum()
c4 = c3.sort_values(ascending=False)
print('\nChecking NA in revised df...')
if sum(c4>0):
sys.exit('\nFAIL: aaindex_df still contains cols with NA, please check and drop these before proceeding...')
else:
print('\nPASS: cols with NA successfully dropped from aaindex_df'
, '\nProceeding with combining aa_df with other features_df')
#---------------------------
# aaindex: sanity check 2
#---------------------------
expected_aa_ncols2 = len(aaindex_df.columns) - sum(c2>0)
if len(aa_df.columns) == expected_aa_ncols2:
print('\nPASS: ncols match'
, '\nExpected ncols:', expected_aa_ncols2
, '\nGot:', len(aa_df.columns))
else:
print('\nFAIL: Numbers mismatch'
, '\nExpected ncols:', expected_aa_ncols2
, '\nGot:', len(aa_df.columns))
# Important: need this to identify aaindex cols
aa_df_cols = aa_df.columns
print('\nTotal no. of columns in clean aa_df:', len(aa_df_cols))
###############################################################################
#%% Combining my_features_df and aaindex_df
#===========================
# Merge my_df + aaindex_df
#===========================
if aa_df.columns[aa_df.columns.isin(my_features_df.columns)] == my_features_df.columns[my_features_df.columns.isin(aa_df.columns)]:
print('\nMerging on column: mutationinformation')
if len(my_features_df) == len(aa_df):
expected_nrows = len(my_features_df)
print('\nProceeding to merge, expected nrows in merged_df:', expected_nrows)
else:
sys.exit('\nNrows mismatch, cannot merge. Please check'
, '\nnrows my_df:', len(my_features_df)
, '\nnrows aa_df:', len(aa_df))
#-----------------
# Reset index: mutationinformation
# Very important for merging
#-----------------
aa_df = aa_df.reset_index()
expected_ncols = len(my_features_df.columns) + len(aa_df.columns) - 1 # for the no. of merging col
#-----------------
# Merge: my_features_df + aa_df
#-----------------
merged_df = pd.merge(my_features_df
, aa_df
, on = 'mutationinformation')
#---------------------------
# aaindex: sanity check 3
#---------------------------
if len(merged_df.columns) == expected_ncols:
print('\nPASS: my_features_df and aa_df successfully combined'
, '\nnrows:', len(merged_df)
, '\nncols:', len(merged_df.columns))
else:
sys.exit('\nFAIL: could not combine my_features_df and aa_df'
, '\nCheck dims and merging cols!')
#--------
# Reassign so downstream code doesn't need to change
#--------
my_df = merged_df.copy()
#%% Data: my_df
# Check if non structural pos have crept in
# IDEALLY remove from source! But for rpoB do it here
# Drop NA where numerical cols have them
if gene.lower() in geneL_na_ppi2:
#D1148 get rid of
na_index = my_df['mutationinformation'].index[my_df['mcsm_na_affinity'].apply(np.isnan)]
my_df = my_df.drop(index=na_index)
# FIXED: complete data for all muts inc L114M, F115L, V123L, V125I, V131M
# if gene.lower() in ['embb']:
# na_index = my_df['mutationinformation'].index[my_df['ligand_distance'].apply(np.isnan)]
# my_df = my_df.drop(index=na_index)
# # Sanity check for non-structural positions
# print('\nChecking for non-structural postions')
# na_index = my_df['mutationinformation'].index[my_df['ligand_distance'].apply(np.isnan)]
# if len(na_index) > 0:
# print('\nNon-structural positions detected for gene:', gene.lower()
# , '\nTotal number of these detected:', len(na_index)
# , '\These are at index:', na_index
# , '\nOriginal nrows:', len(my_df)
# , '\nDropping these...')
# my_df = my_df.drop(index=na_index)
# print('\nRevised nrows:', len(my_df))
# else:
# print('\nNo non-structural positions detected for gene:', gene.lower()
# , '\nnrows:', len(my_df))
###########################################################################
#%% Add lineage calculation columns
#FIXME: Check if this can be imported from config?
total_mtblineage_uc = 8
lineage_colnames = ['lineage_list_all', 'lineage_count_all', 'lineage_count_unique', 'lineage_list_unique', 'lineage_multimode']
#bar = my_df[lineage_colnames]
my_df['lineage_proportion'] = my_df['lineage_count_unique']/my_df['lineage_count_all']
my_df['dist_lineage_proportion'] = my_df['lineage_count_unique']/total_mtblineage_uc
###########################################################################
#%% Active site annotation column
# change from numberic to categorical
if my_df['active_site'].dtype in num_type:
my_df['active_site'] = my_df['active_site'].astype(object)
my_df['active_site'].dtype
#%% AA property change
#--------------------
# Water prop change
#--------------------
my_df['water_change'] = my_df['wt_prop_water'] + str('_to_') + my_df['mut_prop_water']
my_df['water_change'].value_counts()
water_prop_changeD = {
'hydrophobic_to_neutral' : 'change'
, 'hydrophobic_to_hydrophobic' : 'no_change'
, 'neutral_to_neutral' : 'no_change'
, 'neutral_to_hydrophobic' : 'change'
, 'hydrophobic_to_hydrophilic' : 'change'
, 'neutral_to_hydrophilic' : 'change'
, 'hydrophilic_to_neutral' : 'change'
, 'hydrophilic_to_hydrophobic' : 'change'
, 'hydrophilic_to_hydrophilic' : 'no_change'
}
my_df['water_change'] = my_df['water_change'].map(water_prop_changeD)
my_df['water_change'].value_counts()
#--------------------
# Polarity change
#--------------------
my_df['polarity_change'] = my_df['wt_prop_polarity'] + str('_to_') + my_df['mut_prop_polarity']
my_df['polarity_change'].value_counts()
polarity_prop_changeD = {
'non-polar_to_non-polar' : 'no_change'
, 'non-polar_to_neutral' : 'change'
, 'neutral_to_non-polar' : 'change'
, 'neutral_to_neutral' : 'no_change'
, 'non-polar_to_basic' : 'change'
, 'acidic_to_neutral' : 'change'
, 'basic_to_neutral' : 'change'
, 'non-polar_to_acidic' : 'change'
, 'neutral_to_basic' : 'change'
, 'acidic_to_non-polar' : 'change'
, 'basic_to_non-polar' : 'change'
, 'neutral_to_acidic' : 'change'
, 'acidic_to_acidic' : 'no_change'
, 'basic_to_acidic' : 'change'
, 'basic_to_basic' : 'no_change'
, 'acidic_to_basic' : 'change'}
my_df['polarity_change'] = my_df['polarity_change'].map(polarity_prop_changeD)
my_df['polarity_change'].value_counts()
#--------------------
# Electrostatics change
#--------------------
my_df['electrostatics_change'] = my_df['wt_calcprop'] + str('_to_') + my_df['mut_calcprop']
my_df['electrostatics_change'].value_counts()
calc_prop_changeD = {
'non-polar_to_non-polar' : 'no_change'
, 'non-polar_to_polar' : 'change'
, 'polar_to_non-polar' : 'change'
, 'non-polar_to_pos' : 'change'
, 'neg_to_non-polar' : 'change'
, 'non-polar_to_neg' : 'change'
, 'pos_to_polar' : 'change'
, 'pos_to_non-polar' : 'change'
, 'polar_to_polar' : 'no_change'
, 'neg_to_neg' : 'no_change'
, 'polar_to_neg' : 'change'
, 'pos_to_neg' : 'change'
, 'pos_to_pos' : 'no_change'
, 'polar_to_pos' : 'change'
, 'neg_to_polar' : 'change'
, 'neg_to_pos' : 'change'
}
my_df['electrostatics_change'] = my_df['electrostatics_change'].map(calc_prop_changeD)
my_df['electrostatics_change'].value_counts()
#--------------------
# Summary change: Create a combined column summarising these three cols
#--------------------
detect_change = 'change'
check_prop_cols = ['water_change', 'polarity_change', 'electrostatics_change']
#my_df['aa_prop_change'] = (my_df.values == detect_change).any(1).astype(int)
my_df['aa_prop_change'] = (my_df[check_prop_cols].values == detect_change).any(1).astype(int)
my_df['aa_prop_change'].value_counts()
my_df['aa_prop_change'].dtype
my_df['aa_prop_change'] = my_df['aa_prop_change'].map({1:'change'
, 0: 'no_change'})
my_df['aa_prop_change'].value_counts()
my_df['aa_prop_change'].dtype
#%% IMPUTE values for OR [check script for exploration: UQ_or_imputer]
#--------------------
# Impute OR values
#--------------------
#or_cols = ['or_mychisq', 'log10_or_mychisq', 'or_fisher']
sel_cols = ['mutationinformation', 'or_mychisq', 'log10_or_mychisq']
or_cols = ['or_mychisq', 'log10_or_mychisq']
print("count of NULL values before imputation\n")
print(my_df[or_cols].isnull().sum())
my_dfI = pd.DataFrame(index = my_df['mutationinformation'] )
my_dfI = pd.DataFrame(KNN(n_neighbors=3, weights="uniform").fit_transform(my_df[or_cols])
, index = my_df['mutationinformation']
, columns = or_cols )
my_dfI.columns = ['or_rawI', 'logorI']
my_dfI.columns
my_dfI = my_dfI.reset_index(drop = False) # prevents old index from being added as a column
my_dfI.head()
print("count of NULL values AFTER imputation\n")
print(my_dfI.isnull().sum())
#-------------------------------------------
# OR df Merge: with original based on index
#-------------------------------------------
#my_df['index_bm'] = my_df.index
mydf_imputed = pd.merge(my_df
, my_dfI
, on = 'mutationinformation')
#mydf_imputed = mydf_imputed.set_index(['index_bm'])
my_df['log10_or_mychisq'].isna().sum()
mydf_imputed['log10_or_mychisq'].isna().sum()
mydf_imputed['logorI'].isna().sum() # should be 0
len(my_df.columns)
len(mydf_imputed.columns)
#-----------------------------------------
# REASSIGN my_df after imputing OR values
#-----------------------------------------
my_df = mydf_imputed.copy()
if my_df['logorI'].isna().sum() == 0:
print('\nPASS: OR values imputed, data ready for ML')
else:
sys.exit('\nFAIL: something went wrong, Data not ready for ML. Please check upstream!')
#!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
#---------------------------------------
# TODO: try other imputation like MICE
#---------------------------------------
#!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
#%%########################################################################
#==========================
# Data for ML
#==========================
my_df_ml = my_df.copy()
#%% Build X: input for ML
common_cols_stabiltyN = ['ligand_distance'
, 'ligand_affinity_change'
, 'duet_stability_change'
, 'ddg_foldx'
, 'deepddg'
, 'ddg_dynamut2'
, 'mmcsm_lig'
, 'contacts']
# Build stability columns ~ gene
if gene.lower() in geneL_basic:
X_stabilityN = common_cols_stabiltyN
cols_to_mask = ['ligand_affinity_change']
if gene.lower() in geneL_ppi2:
# X_stabilityN = common_cols_stabiltyN + ['mcsm_ppi2_affinity' , 'interface_dist']
geneL_ppi2_st_cols = ['mcsm_ppi2_affinity', 'interface_dist']
X_stabilityN = common_cols_stabiltyN + geneL_ppi2_st_cols
cols_to_mask = ['ligand_affinity_change', 'mcsm_ppi2_affinity']
if gene.lower() in geneL_na:
# X_stabilityN = common_cols_stabiltyN + ['mcsm_na_affinity']
geneL_na_st_cols = ['mcsm_na_affinity']
X_stabilityN = common_cols_stabiltyN + geneL_na_st_cols
cols_to_mask = ['ligand_affinity_change', 'mcsm_na_affinity']
if gene.lower() in geneL_na_ppi2:
# X_stabilityN = common_cols_stabiltyN + ['mcsm_na_affinity'] + ['mcsm_ppi2_affinity', 'interface_dist']
geneL_na_ppi2_st_cols = ['mcsm_na_affinity'] + ['mcsm_ppi2_affinity', 'interface_dist']
X_stabilityN = common_cols_stabiltyN + geneL_na_ppi2_st_cols
cols_to_mask = ['ligand_affinity_change', 'mcsm_na_affinity', 'mcsm_ppi2_affinity']
X_foldX_cols = [ 'electro_rr', 'electro_mm', 'electro_sm', 'electro_ss'
, 'disulfide_rr', 'disulfide_mm', 'disulfide_sm', 'disulfide_ss'
, 'hbonds_rr', 'hbonds_mm', 'hbonds_sm', 'hbonds_ss'
, 'partcov_rr', 'partcov_mm', 'partcov_sm', 'partcov_ss'
, 'vdwclashes_rr', 'vdwclashes_mm', 'vdwclashes_sm', 'vdwclashes_ss'
, 'volumetric_rr', 'volumetric_mm', 'volumetric_ss'
]
X_str = ['rsa'
#, 'asa'
, 'kd_values'
, 'rd_values']
X_ssFN = X_stabilityN + X_str + X_foldX_cols
X_evolFN = ['consurf_score'
, 'snap2_score'
, 'provean_score']
X_genomic_mafor = ['maf'
, 'logorI'
# , 'or_rawI'
# , 'or_mychisq'
# , 'or_logistic'
# , 'or_fisher'
# , 'pval_fisher'
]
X_genomic_linegae = ['lineage_proportion'
, 'dist_lineage_proportion'
#, 'lineage' # could be included as a category but it has L2;L4 formatting
, 'lineage_count_all'
, 'lineage_count_unique'
]
X_genomicFN = X_genomic_mafor + X_genomic_linegae
X_aaindexFN = list(aa_df_cols)
print('\nTotal no. of features for aaindex:', len(X_aaindexFN))
# numerical feature names
numerical_FN = X_ssFN + X_evolFN + X_genomicFN + X_aaindexFN
# categorical feature names
categorical_FN = ['ss_class'
# , 'wt_prop_water'
# , 'mut_prop_water'
# , 'wt_prop_polarity'
# , 'mut_prop_polarity'
# , 'wt_calcprop'
# , 'mut_calcprop'
, 'aa_prop_change'
, 'electrostatics_change'
, 'polarity_change'
, 'water_change'
, 'drtype_mode_labels' # beware then you can't use it to predict [USED it for uq_v1, not v2]
, 'active_site' #[didn't use it for uq_v1]
#, 'gene_name' # will be required for the combined stuff
]
#----------------------------------------------
# count numerical and categorical features
#----------------------------------------------
print('\nNo. of numerical features:', len(numerical_FN)
, '\nNo. of categorical features:', len(categorical_FN))
###########################################################################
#=======================
# Masking columns:
# (mCSM-lig, mCSM-NA, mCSM-ppi2) values for lig_dist >10
#=======================
# my_df_ml['mutationinformation'][my_df['ligand_distance']>10].value_counts()
# my_df_ml.groupby('mutationinformation')['ligand_distance'].apply(lambda x: (x>10)).value_counts()
# my_df_ml.loc[(my_df_ml['ligand_distance'] > 10), 'ligand_affinity_change'] = 0
# (my_df_ml['ligand_affinity_change'] == 0).sum()
my_df_ml['mutationinformation'][my_df_ml['ligand_distance']>10].value_counts()
my_df_ml.groupby('mutationinformation')['ligand_distance'].apply(lambda x: (x>10)).value_counts()
my_df_ml.loc[(my_df_ml['ligand_distance'] > 10), cols_to_mask].value_counts()
# mask the mcsm affinity related columns where ligand distance > 10
my_df_ml.loc[(my_df_ml['ligand_distance'] > 10), cols_to_mask] = 0
(my_df_ml['ligand_affinity_change'] == 0).sum()
mask_check = my_df_ml[['mutationinformation', 'ligand_distance'] + cols_to_mask]
# write file for check
mask_check.sort_values(by = ['ligand_distance'], ascending = True, inplace = True)
mask_check.to_csv(outdir + 'ml/' + gene.lower() + '_mask_check.csv')
#=================================================
# Training and BLIND test set: imputed vs actual
# BUT in REVERSE i.e.
# dst with actual values : blind test
# dst with imputed values : training set
#==================================================
my_df_ml[drug].isna().sum() #'na' ones are now training set
blind_test_df = my_df_ml[my_df_ml[drug].notna()]
blind_test_df.shape
training_df = my_df_ml[my_df_ml[drug].isna()]
training_df.shape
# Target 1: dst_mode
training_df[drug].value_counts()
training_df['dst_mode'].value_counts()
####################################################################
#%% extracting dfs based on numerical, categorical column names
#----------------------------------
# WITHOUT the target var included
#----------------------------------
num_df = training_df[numerical_FN]
num_df.shape
cat_df = training_df[categorical_FN]
cat_df.shape
all_df = training_df[numerical_FN + categorical_FN]
all_df.shape
#------------------------------
# WITH the target var included:
#'wtgt': with target
#------------------------------
# drug and dst_mode should be the same thing
num_df_wtgt = training_df[numerical_FN + ['dst_mode']]
num_df_wtgt.shape
cat_df_wtgt = training_df[categorical_FN + ['dst_mode']]
cat_df_wtgt.shape
all_df_wtgt = training_df[numerical_FN + categorical_FN + ['dst_mode']]
all_df_wtgt.shape
#%%########################################################################
#============
# ML data
#============
#------
# X: Training and Blind test (BTS)
#------
X = all_df_wtgt[numerical_FN + categorical_FN] # training data ALL
X_bts = blind_test_df[numerical_FN + categorical_FN] # blind test data ALL
#X = all_df_wtgt[numerical_FN] # training numerical only
#X_bts = blind_test_df[numerical_FN] # blind test data numerical
#------
# y
#------
y = all_df_wtgt['dst_mode'] # training data y
y_bts = blind_test_df['dst_mode'] # blind data test y
#X_bts_wt = blind_test_df[numerical_FN + ['dst_mode']]
# Quick check
#(X['ligand_affinity_change']==0).sum() == (X['ligand_distance']>10).sum()
for i in range(len(cols_to_mask)):
ind = i+1
print('\nindex:', i, '\nind:', ind)
print('\nMask count check:'
, (my_df_ml[cols_to_mask[i]]==0).sum() == (my_df_ml['ligand_distance']>10).sum()
)
print('Original Data\n', Counter(y)
, 'Data dim:', X.shape)
yc1 = Counter(y)
yc1_ratio = yc1[0]/yc1[1]
yc2 = Counter(y_bts)
yc2_ratio = yc2[0]/yc2[1]
print('\n-------------------------------------------------------------'
, '\nSuccessfully split data: REVERSE training'
, '\nimputed values: training set'
, '\nactual values: blind test set'
, '\nTrain data size:', X.shape
, '\nTest data size:', X_bts.shape
, '\ny_train numbers:', yc1
, '\ny_train ratio:',yc1_ratio
, '\n'
, '\ny_test_numbers:', yc2
, '\ny_test ratio:', yc2_ratio
, '\n-------------------------------------------------------------'
)
###########################################################################
#%%
###########################################################################
# RESAMPLING
###########################################################################
#------------------------------
# Simple Random oversampling
# [Numerical + catgeorical]
#------------------------------
oversample = RandomOverSampler(sampling_strategy='minority')
X_ros, y_ros = oversample.fit_resample(X, y)
print('Simple Random OverSampling\n', Counter(y_ros))
print(X_ros.shape)
#------------------------------
# Simple Random Undersampling
# [Numerical + catgeorical]
#------------------------------
undersample = RandomUnderSampler(sampling_strategy='majority')
X_rus, y_rus = undersample.fit_resample(X, y)
print('Simple Random UnderSampling\n', Counter(y_rus))
print(X_rus.shape)
#------------------------------
# Simple combine ROS and RUS
# [Numerical + catgeorical]
#------------------------------
oversample = RandomOverSampler(sampling_strategy='minority')
X_ros, y_ros = oversample.fit_resample(X, y)
undersample = RandomUnderSampler(sampling_strategy='majority')
X_rouC, y_rouC = undersample.fit_resample(X_ros, y_ros)
print('Simple Combined Over and UnderSampling\n', Counter(y_rouC))
print(X_rouC.shape)
#------------------------------
# SMOTE_NC: oversampling
# [numerical + categorical]
#https://stackoverflow.com/questions/47655813/oversampling-smote-for-binary-and-categorical-data-in-python
#------------------------------
# Determine categorical and numerical features
numerical_ix = X.select_dtypes(include=['int64', 'float64']).columns
numerical_ix
num_featuresL = list(numerical_ix)
numerical_colind = X.columns.get_indexer(list(numerical_ix) )
numerical_colind
categorical_ix = X.select_dtypes(include=['object', 'bool']).columns
categorical_ix
categorical_colind = X.columns.get_indexer(list(categorical_ix))
categorical_colind
#k_sm = 5 #default, but this fails for gid as n_samples 3 [ONLY for reverse training]
if gene.lower() in geneL_na:
k_sm = 1
else:
k_sm = 5
sm_nc = SMOTENC(categorical_features=categorical_colind, k_neighbors = k_sm, **rs, **njobs)
X_smnc, y_smnc = sm_nc.fit_resample(X, y)
print('SMOTE_NC OverSampling\n', Counter(y_smnc))
print(X_smnc.shape)
globals().update(locals()) # TROLOLOLOLOLOLS
#print("i did a horrible hack :-)")
###############################################################################
#%% SMOTE RESAMPLING for NUMERICAL ONLY*
# #------------------------------
# # SMOTE: Oversampling
# # [Numerical ONLY]
# #------------------------------
# k_sm = 1
# sm = SMOTE(sampling_strategy = 'auto', k_neighbors = k_sm, **rs)
# X_sm, y_sm = sm.fit_resample(X, y)
# print(X_sm.shape)
# print('SMOTE OverSampling\n', Counter(y_sm))
# y_sm_df = y_sm.to_frame()
# y_sm_df.value_counts().plot(kind = 'bar')
# #------------------------------
# # SMOTE: Over + Undersampling COMBINED
# # [Numerical ONLY]
# #-----------------------------
# sm_enn = SMOTEENN(enn=EditedNearestNeighbours(sampling_strategy='all', **rs, **njobs ))
# X_enn, y_enn = sm_enn.fit_resample(X, y)
# print(X_enn.shape)
# print('SMOTE Over+Under Sampling combined\n', Counter(y_enn))
###############################################################################
# TODO: Find over and undersampling JUST for categorical data

811
scripts/ml/ml_data_sl.py
View file

@ -1,811 +0,0 @@
#!/usr/bin/env python3
# -*- coding: utf-8 -*-
"""
Created on Sun Mar 6 13:41:54 2022
@author: tanu
"""
def setvars(gene,drug):
#https://stackoverflow.com/questions/51695322/compare-multiple-algorithms-with-sklearn-pipeline
import os, sys
import pandas as pd
import numpy as np
print(np.__version__)
print(pd.__version__)
import pprint as pp
from copy import deepcopy
from collections import Counter
from sklearn.impute import KNNImputer as KNN
from imblearn.over_sampling import RandomOverSampler
from imblearn.under_sampling import RandomUnderSampler
from imblearn.over_sampling import SMOTE
from sklearn.datasets import make_classification
from imblearn.combine import SMOTEENN
from imblearn.combine import SMOTETomek
from imblearn.over_sampling import SMOTENC
from imblearn.under_sampling import EditedNearestNeighbours
from imblearn.under_sampling import RepeatedEditedNearestNeighbours
from sklearn.metrics import make_scorer, confusion_matrix, accuracy_score, balanced_accuracy_score, precision_score, average_precision_score, recall_score
from sklearn.metrics import roc_auc_score, roc_curve, f1_score, matthews_corrcoef, jaccard_score, classification_report
from sklearn.model_selection import train_test_split, cross_validate, cross_val_score
from sklearn.model_selection import StratifiedKFold,RepeatedStratifiedKFold, RepeatedKFold
from sklearn.pipeline import Pipeline, make_pipeline
import argparse
import re
#%% GLOBALS
tts_split = "sl"
rs = {'random_state': 42}
njobs = {'n_jobs': 10}
scoring_fn = ({ 'mcc' : make_scorer(matthews_corrcoef)
, 'accuracy' : make_scorer(accuracy_score)
, 'fscore' : make_scorer(f1_score)
, 'precision' : make_scorer(precision_score)
, 'recall' : make_scorer(recall_score)
, 'roc_auc' : make_scorer(roc_auc_score)
, 'jcc' : make_scorer(jaccard_score)
})
skf_cv = StratifiedKFold(n_splits = 10
#, shuffle = False, random_state= None)
, shuffle = True,**rs)
rskf_cv = RepeatedStratifiedKFold(n_splits = 10
, n_repeats = 3
, **rs)
mcc_score_fn = {'mcc': make_scorer(matthews_corrcoef)}
jacc_score_fn = {'jcc': make_scorer(jaccard_score)}
#%% FOR LATER: Combine ED logo data
###########################################################################
homedir = os.path.expanduser("~")
geneL_basic = ['pnca']
geneL_na = ['gid']
geneL_na_ppi2 = ['rpob']
geneL_ppi2 = ['alr', 'embb', 'katg']
#num_type = ['int64', 'float64']
num_type = ['int16', 'int32', 'int64', 'float16', 'float32', 'float64']
cat_type = ['object', 'bool']
#==============
# directories
#==============
datadir = homedir + '/git/Data/'
indir = datadir + drug + '/input/'
outdir = datadir + drug + '/output/'
#=======
# input
#=======
#---------
# File 1
#---------
infile_ml1 = outdir + gene.lower() + '_merged_df3.csv'
#infile_ml2 = outdir + gene.lower() + '_merged_df2.csv'
my_features_df = pd.read_csv(infile_ml1, index_col = 0)
my_features_df = my_features_df .reset_index(drop = True)
my_features_df.index
my_features_df.dtypes
mycols = my_features_df.columns
#---------
# File 2
#---------
infile_aaindex = outdir + 'aa_index/' + gene.lower() + '_aa.csv'
aaindex_df = pd.read_csv(infile_aaindex, index_col = 0)
aaindex_df.dtypes
#-----------
# check for non-numerical columns
#-----------
if any(aaindex_df.dtypes==object):
print('\naaindex_df contains non-numerical data')
aaindex_df_object = aaindex_df.select_dtypes(include = cat_type)
print('\nTotal no. of non-numerial columns:', len(aaindex_df_object.columns))
expected_aa_ncols = len(aaindex_df.columns) - len(aaindex_df_object.columns)
#-----------
# Extract numerical data only
#-----------
print('\nSelecting numerical data only')
aaindex_df = aaindex_df.select_dtypes(include = num_type)
#---------------------------
# aaindex: sanity check 1
#---------------------------
if len(aaindex_df.columns) == expected_aa_ncols:
print('\nPASS: successfully selected numerical columns only for aaindex_df')
else:
print('\nFAIL: Numbers mismatch'
, '\nExpected ncols:', expected_aa_ncols
, '\nGot:', len(aaindex_df.columns))
#---------------
# check for NA
#---------------
print('\nNow checking for NA in the remaining aaindex_cols')
c1 = aaindex_df.isna().sum()
c2 = c1.sort_values(ascending=False)
print('\nCounting aaindex_df cols with NA'
, '\nncols with NA:', sum(c2>0), 'columns'
, '\nDropping these...'
, '\nOriginal ncols:', len(aaindex_df.columns)
)
aa_df = aaindex_df.dropna(axis=1)
print('\nRevised df ncols:', len(aa_df.columns))
c3 = aa_df.isna().sum()
c4 = c3.sort_values(ascending=False)
print('\nChecking NA in revised df...')
if sum(c4>0):
sys.exit('\nFAIL: aaindex_df still contains cols with NA, please check and drop these before proceeding...')
else:
print('\nPASS: cols with NA successfully dropped from aaindex_df'
, '\nProceeding with combining aa_df with other features_df')
#---------------------------
# aaindex: sanity check 2
#---------------------------
expected_aa_ncols2 = len(aaindex_df.columns) - sum(c2>0)
if len(aa_df.columns) == expected_aa_ncols2:
print('\nPASS: ncols match'
, '\nExpected ncols:', expected_aa_ncols2
, '\nGot:', len(aa_df.columns))
else:
print('\nFAIL: Numbers mismatch'
, '\nExpected ncols:', expected_aa_ncols2
, '\nGot:', len(aa_df.columns))
# Important: need this to identify aaindex cols
aa_df_cols = aa_df.columns
print('\nTotal no. of columns in clean aa_df:', len(aa_df_cols))
###############################################################################
#%% Combining my_features_df and aaindex_df
#===========================
# Merge my_df + aaindex_df
#===========================
if aa_df.columns[aa_df.columns.isin(my_features_df.columns)] == my_features_df.columns[my_features_df.columns.isin(aa_df.columns)]:
print('\nMerging on column: mutationinformation')
if len(my_features_df) == len(aa_df):
expected_nrows = len(my_features_df)
print('\nProceeding to merge, expected nrows in merged_df:', expected_nrows)
else:
sys.exit('\nNrows mismatch, cannot merge. Please check'
, '\nnrows my_df:', len(my_features_df)
, '\nnrows aa_df:', len(aa_df))
#-----------------
# Reset index: mutationinformation
# Very important for merging
#-----------------
aa_df = aa_df.reset_index()
expected_ncols = len(my_features_df.columns) + len(aa_df.columns) - 1 # for the no. of merging col
#-----------------
# Merge: my_features_df + aa_df
#-----------------
merged_df = pd.merge(my_features_df
, aa_df
, on = 'mutationinformation')
#---------------------------
# aaindex: sanity check 3
#---------------------------
if len(merged_df.columns) == expected_ncols:
print('\nPASS: my_features_df and aa_df successfully combined'
, '\nnrows:', len(merged_df)
, '\nncols:', len(merged_df.columns))
else:
sys.exit('\nFAIL: could not combine my_features_df and aa_df'
, '\nCheck dims and merging cols!')
#--------
# Reassign so downstream code doesn't need to change
#--------
my_df = merged_df.copy()
#%% Data: my_df
# Check if non structural pos have crept in
# IDEALLY remove from source! But for rpoB do it here
# Drop NA where numerical cols have them
if gene.lower() in geneL_na_ppi2:
#D1148 get rid of
na_index = my_df['mutationinformation'].index[my_df['mcsm_na_affinity'].apply(np.isnan)]
my_df = my_df.drop(index=na_index)
# FIXED: complete data for all muts inc L114M, F115L, V123L, V125I, V131M
# if gene.lower() in ['embb']:
# na_index = my_df['mutationinformation'].index[my_df['ligand_distance'].apply(np.isnan)]
# my_df = my_df.drop(index=na_index)
# # Sanity check for non-structural positions
# print('\nChecking for non-structural postions')
# na_index = my_df['mutationinformation'].index[my_df['ligand_distance'].apply(np.isnan)]
# if len(na_index) > 0:
# print('\nNon-structural positions detected for gene:', gene.lower()
# , '\nTotal number of these detected:', len(na_index)
# , '\These are at index:', na_index
# , '\nOriginal nrows:', len(my_df)
# , '\nDropping these...')
# my_df = my_df.drop(index=na_index)
# print('\nRevised nrows:', len(my_df))
# else:
# print('\nNo non-structural positions detected for gene:', gene.lower()
# , '\nnrows:', len(my_df))
###########################################################################
#%% Add lineage calculation columns
#FIXME: Check if this can be imported from config?
total_mtblineage_uc = 8
lineage_colnames = ['lineage_list_all', 'lineage_count_all', 'lineage_count_unique', 'lineage_list_unique', 'lineage_multimode']
#bar = my_df[lineage_colnames]
my_df['lineage_proportion'] = my_df['lineage_count_unique']/my_df['lineage_count_all']
my_df['dist_lineage_proportion'] = my_df['lineage_count_unique']/total_mtblineage_uc
###########################################################################
#%% Active site annotation column
# change from numberic to categorical
if my_df['active_site'].dtype in num_type:
my_df['active_site'] = my_df['active_site'].astype(object)
my_df['active_site'].dtype
#%% AA property change
#--------------------
# Water prop change
#--------------------
my_df['water_change'] = my_df['wt_prop_water'] + str('_to_') + my_df['mut_prop_water']
my_df['water_change'].value_counts()
water_prop_changeD = {
'hydrophobic_to_neutral' : 'change'
, 'hydrophobic_to_hydrophobic' : 'no_change'
, 'neutral_to_neutral' : 'no_change'
, 'neutral_to_hydrophobic' : 'change'
, 'hydrophobic_to_hydrophilic' : 'change'
, 'neutral_to_hydrophilic' : 'change'
, 'hydrophilic_to_neutral' : 'change'
, 'hydrophilic_to_hydrophobic' : 'change'
, 'hydrophilic_to_hydrophilic' : 'no_change'
}
my_df['water_change'] = my_df['water_change'].map(water_prop_changeD)
my_df['water_change'].value_counts()
#--------------------
# Polarity change
#--------------------
my_df['polarity_change'] = my_df['wt_prop_polarity'] + str('_to_') + my_df['mut_prop_polarity']
my_df['polarity_change'].value_counts()
polarity_prop_changeD = {
'non-polar_to_non-polar' : 'no_change'
, 'non-polar_to_neutral' : 'change'
, 'neutral_to_non-polar' : 'change'
, 'neutral_to_neutral' : 'no_change'
, 'non-polar_to_basic' : 'change'
, 'acidic_to_neutral' : 'change'
, 'basic_to_neutral' : 'change'
, 'non-polar_to_acidic' : 'change'
, 'neutral_to_basic' : 'change'
, 'acidic_to_non-polar' : 'change'
, 'basic_to_non-polar' : 'change'
, 'neutral_to_acidic' : 'change'
, 'acidic_to_acidic' : 'no_change'
, 'basic_to_acidic' : 'change'
, 'basic_to_basic' : 'no_change'
, 'acidic_to_basic' : 'change'}
my_df['polarity_change'] = my_df['polarity_change'].map(polarity_prop_changeD)
my_df['polarity_change'].value_counts()
#--------------------
# Electrostatics change
#--------------------
my_df['electrostatics_change'] = my_df['wt_calcprop'] + str('_to_') + my_df['mut_calcprop']
my_df['electrostatics_change'].value_counts()
calc_prop_changeD = {
'non-polar_to_non-polar' : 'no_change'
, 'non-polar_to_polar' : 'change'
, 'polar_to_non-polar' : 'change'
, 'non-polar_to_pos' : 'change'
, 'neg_to_non-polar' : 'change'
, 'non-polar_to_neg' : 'change'
, 'pos_to_polar' : 'change'
, 'pos_to_non-polar' : 'change'
, 'polar_to_polar' : 'no_change'
, 'neg_to_neg' : 'no_change'
, 'polar_to_neg' : 'change'
, 'pos_to_neg' : 'change'
, 'pos_to_pos' : 'no_change'
, 'polar_to_pos' : 'change'
, 'neg_to_polar' : 'change'
, 'neg_to_pos' : 'change'
}
my_df['electrostatics_change'] = my_df['electrostatics_change'].map(calc_prop_changeD)
my_df['electrostatics_change'].value_counts()
#--------------------
# Summary change: Create a combined column summarising these three cols
#--------------------
detect_change = 'change'
check_prop_cols = ['water_change', 'polarity_change', 'electrostatics_change']
#my_df['aa_prop_change'] = (my_df.values == detect_change).any(1).astype(int)
my_df['aa_prop_change'] = (my_df[check_prop_cols].values == detect_change).any(1).astype(int)
my_df['aa_prop_change'].value_counts()
my_df['aa_prop_change'].dtype
my_df['aa_prop_change'] = my_df['aa_prop_change'].map({1:'change'
, 0: 'no_change'})
my_df['aa_prop_change'].value_counts()
my_df['aa_prop_change'].dtype
#%% IMPUTE values for OR [check script for exploration: UQ_or_imputer]
#--------------------
# Impute OR values
#--------------------
#or_cols = ['or_mychisq', 'log10_or_mychisq', 'or_fisher']
sel_cols = ['mutationinformation', 'or_mychisq', 'log10_or_mychisq']
or_cols = ['or_mychisq', 'log10_or_mychisq']
print("count of NULL values before imputation\n")
print(my_df[or_cols].isnull().sum())
my_dfI = pd.DataFrame(index = my_df['mutationinformation'] )
my_dfI = pd.DataFrame(KNN(n_neighbors=3, weights="uniform").fit_transform(my_df[or_cols])
, index = my_df['mutationinformation']
, columns = or_cols )
my_dfI.columns = ['or_rawI', 'logorI']
my_dfI.columns
my_dfI = my_dfI.reset_index(drop = False) # prevents old index from being added as a column
my_dfI.head()
print("count of NULL values AFTER imputation\n")
print(my_dfI.isnull().sum())
#-------------------------------------------
# OR df Merge: with original based on index
#-------------------------------------------
#my_df['index_bm'] = my_df.index
mydf_imputed = pd.merge(my_df
, my_dfI
, on = 'mutationinformation')
#mydf_imputed = mydf_imputed.set_index(['index_bm'])
my_df['log10_or_mychisq'].isna().sum()
mydf_imputed['log10_or_mychisq'].isna().sum()
mydf_imputed['logorI'].isna().sum() # should be 0
len(my_df.columns)
len(mydf_imputed.columns)
#-----------------------------------------
# REASSIGN my_df after imputing OR values
#-----------------------------------------
my_df = mydf_imputed.copy()
if my_df['logorI'].isna().sum() == 0:
print('\nPASS: OR values imputed, data ready for ML')
else:
sys.exit('\nFAIL: something went wrong, Data not ready for ML. Please check upstream!')
#!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
#---------------------------------------
# TODO: try other imputation like MICE
#---------------------------------------
#!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
#%%########################################################################
#==========================
# Data for ML
#==========================
my_df_ml = my_df.copy()
# Build column names to mask for affinity chanhes
if gene.lower() in geneL_basic:
#X_stabilityN = common_cols_stabiltyN
gene_affinity_colnames = []# not needed as its the common ones
cols_to_mask = ['ligand_affinity_change']
if gene.lower() in geneL_ppi2:
gene_affinity_colnames = ['mcsm_ppi2_affinity', 'interface_dist']
#X_stabilityN = common_cols_stabiltyN + geneL_ppi2_st_cols
cols_to_mask = ['ligand_affinity_change', 'mcsm_ppi2_affinity']
if gene.lower() in geneL_na:
gene_affinity_colnames = ['mcsm_na_affinity']
#X_stabilityN = common_cols_stabiltyN + geneL_na_st_cols
cols_to_mask = ['ligand_affinity_change', 'mcsm_na_affinity']
if gene.lower() in geneL_na_ppi2:
gene_affinity_colnames = ['mcsm_na_affinity'] + ['mcsm_ppi2_affinity', 'interface_dist']
#X_stabilityN = common_cols_stabiltyN + geneL_na_ppi2_st_cols
cols_to_mask = ['ligand_affinity_change', 'mcsm_na_affinity', 'mcsm_ppi2_affinity']
#=======================
# Masking columns:
# (mCSM-lig, mCSM-NA, mCSM-ppi2) values for lig_dist >10
#=======================
my_df_ml['mutationinformation'][my_df_ml['ligand_distance']>10].value_counts()
my_df_ml.groupby('mutationinformation')['ligand_distance'].apply(lambda x: (x>10)).value_counts()
my_df_ml.loc[(my_df_ml['ligand_distance'] > 10), cols_to_mask].value_counts()
# mask the mcsm affinity related columns where ligand distance > 10
my_df_ml.loc[(my_df_ml['ligand_distance'] > 10), cols_to_mask] = 0
(my_df_ml['ligand_affinity_change'] == 0).sum()
mask_check = my_df_ml[['mutationinformation', 'ligand_distance'] + cols_to_mask]
#===================================================
# write file for check
mask_check.sort_values(by = ['ligand_distance'], ascending = True, inplace = True)
mask_check.to_csv(outdir + 'ml/' + gene.lower() + '_mask_check.csv')
#===================================================
###############################################################################
#%% Feature groups (FG): Build X for Input ML
############################################################################
#===========================
# FG1: Evolutionary features
#===========================
X_evolFN = ['consurf_score'
, 'snap2_score'
, 'provean_score']
###############################################################################
#========================
# FG2: Stability features
#========================
#--------
# common
#--------
X_common_stability_Fnum = [
'duet_stability_change'
, 'ddg_foldx'
, 'deepddg'
, 'ddg_dynamut2'
, 'contacts']
#--------
# FoldX
#--------
X_foldX_Fnum = [ 'electro_rr', 'electro_mm', 'electro_sm', 'electro_ss'
, 'disulfide_rr', 'disulfide_mm', 'disulfide_sm', 'disulfide_ss'
, 'hbonds_rr', 'hbonds_mm', 'hbonds_sm', 'hbonds_ss'
, 'partcov_rr', 'partcov_mm', 'partcov_sm', 'partcov_ss'
, 'vdwclashes_rr', 'vdwclashes_mm', 'vdwclashes_sm', 'vdwclashes_ss'
, 'volumetric_rr', 'volumetric_mm', 'volumetric_ss']
X_stability_FN = X_common_stability_Fnum + X_foldX_Fnum
###############################################################################
#===================
# FG3: Affinity features
#===================
common_affinity_Fnum = ['ligand_distance'
, 'ligand_affinity_change'
, 'mmcsm_lig']
# if gene.lower() in geneL_basic:
# X_affinityFN = common_affinity_Fnum
# else:
# X_affinityFN = common_affinity_Fnum + gene_affinity_colnames
X_affinityFN = common_affinity_Fnum + gene_affinity_colnames
###############################################################################
#============================
# FG4: Residue level features
#============================
#-----------
# AA index
#-----------
X_aaindex_Fnum = list(aa_df_cols)
print('\nTotal no. of features for aaindex:', len(X_aaindex_Fnum))
#-----------------
# surface area
# depth
# hydrophobicity
#-----------------
X_str_Fnum = ['rsa'
#, 'asa'
, 'kd_values'
, 'rd_values']
#---------------------------
# Other aa properties
# active site indication
#---------------------------
X_aap_Fcat = ['ss_class'
# , 'wt_prop_water'
# , 'mut_prop_water'
# , 'wt_prop_polarity'
# , 'mut_prop_polarity'
# , 'wt_calcprop'
# , 'mut_calcprop'
, 'aa_prop_change'
, 'electrostatics_change'
, 'polarity_change'
, 'water_change'
, 'active_site']
X_resprop_FN = X_aaindex_Fnum + X_str_Fnum + X_aap_Fcat
###############################################################################
#========================
# FG5: Genomic features
#========================
X_gn_mafor_Fnum = ['maf'
#, 'logorI'
# , 'or_rawI'
# , 'or_mychisq'
# , 'or_logistic'
# , 'or_fisher'
# , 'pval_fisher'
]
X_gn_linegae_Fnum = ['lineage_proportion'
, 'dist_lineage_proportion'
#, 'lineage' # could be included as a category but it has L2;L4 formatting
, 'lineage_count_all'
, 'lineage_count_unique'
]
# X_gn_Fcat = ['drtype_mode_labels' # beware then you can't use it to predict [USED it for uq_v1, not v2]
# #, 'gene_name' # will be required for the combined stuff
# ]
X_gn_Fcat = []
X_genomicFN = X_gn_mafor_Fnum + X_gn_linegae_Fnum + X_gn_Fcat
###############################################################################
#========================
# FG6 collapsed: Structural : Atability + Affinity + ResidueProp
#========================
X_structural_FN = X_stability_FN + X_affinityFN + X_resprop_FN
###############################################################################
#========================
# BUILDING all features
#========================
all_featuresN = X_evolFN + X_structural_FN + X_genomicFN
###############################################################################
#%% Define training and test data
#================================================================
# Training and BLIND test set: scaling law split
#https://towardsdatascience.com/finally-why-we-use-an-80-20-split-for-training-and-test-data-plus-an-alternative-method-oh-yes-edc77e96295d
# dst with actual values : training set
# dst with imputed values : THROW AWAY [unrepresentative]
# test data size ~ 1/sqrt(features NOT including target variable)
#================================================================
my_df_ml[drug].isna().sum()
# blind_test_df = my_df_ml[my_df_ml[drug].isna()]
# blind_test_df.shape
training_df = my_df_ml[my_df_ml[drug].notna()]
training_df.shape
# Target 1: dst_mode
training_df[drug].value_counts()
training_df['dst_mode'].value_counts()
####################################################################
#====================================
# ML data: Train test split: SL
# with stratification
# 1-blind test : training_data for CV
# 1/sqrt(columns) : blind test
#===========================================
x_features = training_df[all_featuresN]
y_target = training_df['dst_mode']
# sanity check
if not 'dst_mode' in x_features.columns:
print('\nPASS: x_features has no target variable')
x_ncols = len(x_features.columns)
print('\nNo. of columns for x_features:', x_ncols)
# NEED It for scaling law split
#https://towardsdatascience.com/finally-why-we-use-an-80-20-split-for-training-and-test-data-plus-an-alternative-method-oh-yes-edc77e96295d
else:
sys.exit('\nFAIL: x_features has target variable included. FIX it and rerun!')
#-------------------
# train-test split
#-------------------
sl_test_size = 1/np.sqrt(x_ncols)
train = 1 - sl_test_size
#x_train, x_test, y_train, y_test # traditional var_names
# so my downstream code doesn't need to change
X, X_bts, y, y_bts = train_test_split(x_features, y_target
, test_size = sl_test_size
, **rs
, stratify = y_target)
yc1 = Counter(y)
yc1_ratio = yc1[0]/yc1[1]
yc2 = Counter(y_bts)
yc2_ratio = yc2[0]/yc2[1]
###############################################################################
#======================================================
# Determine categorical and numerical features
#======================================================
numerical_cols = X.select_dtypes(include=['int64', 'float64']).columns
numerical_cols
categorical_cols = X.select_dtypes(include=['object', 'bool']).columns
categorical_cols
################################################################################
# IMPORTANT sanity checks
if len(X.columns) == len(X_evolFN) + len(X_stability_FN) + len(X_affinityFN) + len(X_resprop_FN) + len(X_genomicFN):
print('\nPASS: ML data with input features, training and test generated...'
, '\n\nTotal no. of input features:' , len(X.columns)
, '\n--------No. of numerical features:' , len(numerical_cols)
, '\n--------No. of categorical features:' , len(categorical_cols)
, '\n\nTotal no. of evolutionary features:' , len(X_evolFN)
, '\n\nTotal no. of stability features:' , len(X_stability_FN)
, '\n--------Common stabilty cols:' , len(X_common_stability_Fnum)
, '\n--------Foldx cols:' , len(X_foldX_Fnum)
, '\n\nTotal no. of affinity features:' , len(X_affinityFN)
, '\n--------Common affinity cols:' , len(common_affinity_Fnum)
, '\n--------Gene specific affinity cols:' , len(gene_affinity_colnames)
, '\n\nTotal no. of residue level features:', len(X_resprop_FN)
, '\n--------AA index cols:' , len(X_aaindex_Fnum)
, '\n--------Residue Prop cols:' , len(X_str_Fnum)
, '\n--------AA change Prop cols:' , len(X_aap_Fcat)
, '\n\nTotal no. of genomic features:' , len(X_genomicFN)
, '\n--------MAF+OR cols:' , len(X_gn_mafor_Fnum)
, '\n--------Lineage cols:' , len(X_gn_linegae_Fnum)
, '\n--------Other cols:' , len(X_gn_Fcat)
)
else:
print('\nFAIL: numbers mismatch'
, '\nExpected:',len(X_evolFN) + len(X_stability_FN) + len(X_affinityFN) + len(X_resprop_FN) + len(X_genomicFN)
, '\nGot:', len(X.columns))
sys.exit()
###############################################################################
print('\n-------------------------------------------------------------'
, '\nSuccessfully split data: ALL features'
, '\nactual values: training set'
, '\nSplit:', tts_split
#, '\nimputed values: blind test set'
, '\n\nTotal data size:', len(X) + len(X_bts)
, '\n\nTrain data size:', X.shape
, '\ny_train numbers:', yc1
, '\n\nTest data size:', X_bts.shape
, '\ny_test_numbers:', yc2
, '\n\ny_train ratio:',yc1_ratio
, '\ny_test ratio:', yc2_ratio
, '\n-------------------------------------------------------------'
)
##########################################################################
# Quick check
#(X['ligand_affinity_change']==0).sum() == (X['ligand_distance']>10).sum()
for i in range(len(cols_to_mask)):
ind = i+1
print('\nindex:', i, '\nind:', ind)
print('\nMask count check:'
, (my_df_ml[cols_to_mask[i]]==0).sum() == (my_df_ml['ligand_distance']>10).sum()
)
print('Original Data\n', Counter(y)
, 'Data dim:', X.shape)
###########################################################################
#%%
###########################################################################
# RESAMPLING
###########################################################################
#------------------------------
# Simple Random oversampling
# [Numerical + catgeorical]
#------------------------------
oversample = RandomOverSampler(sampling_strategy='minority')
X_ros, y_ros = oversample.fit_resample(X, y)
print('\nSimple Random OverSampling\n', Counter(y_ros))
print(X_ros.shape)
#------------------------------
# Simple Random Undersampling
# [Numerical + catgeorical]
#------------------------------
undersample = RandomUnderSampler(sampling_strategy='majority')
X_rus, y_rus = undersample.fit_resample(X, y)
print('\nSimple Random UnderSampling\n', Counter(y_rus))
print(X_rus.shape)
#------------------------------
# Simple combine ROS and RUS
# [Numerical + catgeorical]
#------------------------------
oversample = RandomOverSampler(sampling_strategy='minority')
X_ros, y_ros = oversample.fit_resample(X, y)
undersample = RandomUnderSampler(sampling_strategy='majority')
X_rouC, y_rouC = undersample.fit_resample(X_ros, y_ros)
print('\nSimple Combined Over and UnderSampling\n', Counter(y_rouC))
print(X_rouC.shape)
#------------------------------
# SMOTE_NC: oversampling
# [numerical + categorical]
#https://stackoverflow.com/questions/47655813/oversampling-smote-for-binary-and-categorical-data-in-python
#------------------------------
# Determine categorical and numerical features
numerical_ix = X.select_dtypes(include=['int64', 'float64']).columns
numerical_ix
num_featuresL = list(numerical_ix)
numerical_colind = X.columns.get_indexer(list(numerical_ix) )
numerical_colind
categorical_ix = X.select_dtypes(include=['object', 'bool']).columns
categorical_ix
categorical_colind = X.columns.get_indexer(list(categorical_ix))
categorical_colind
k_sm = 5 # 5 is default
sm_nc = SMOTENC(categorical_features=categorical_colind, k_neighbors = k_sm, **rs, **njobs)
X_smnc, y_smnc = sm_nc.fit_resample(X, y)
print('\nSMOTE_NC OverSampling\n', Counter(y_smnc))
print(X_smnc.shape)
globals().update(locals()) # TROLOLOLOLOLOLS
#print("i did a horrible hack :-)")
###############################################################################
#%% SMOTE RESAMPLING for NUMERICAL ONLY*
# #------------------------------
# # SMOTE: Oversampling
# # [Numerical ONLY]
# #------------------------------
# k_sm = 1
# sm = SMOTE(sampling_strategy = 'auto', k_neighbors = k_sm, **rs)
# X_sm, y_sm = sm.fit_resample(X, y)
# print(X_sm.shape)
# print('\nSMOTE OverSampling\n', Counter(y_sm))
# y_sm_df = y_sm.to_frame()
# y_sm_df.value_counts().plot(kind = 'bar')
# #------------------------------
# # SMOTE: Over + Undersampling COMBINED
# # [Numerical ONLY]
# #-----------------------------
# sm_enn = SMOTEENN(enn=EditedNearestNeighbours(sampling_strategy='all', **rs, **njobs ))
# X_enn, y_enn = sm_enn.fit_resample(X, y)
# print(X_enn.shape)
# print('\nSMOTE Over+Under Sampling combined\n', Counter(y_enn))
###############################################################################
# TODO: Find over and undersampling JUST for categorical data
###########################################################################
print('\n#################################################################'
, '\nDim of X for gene:', gene.lower(), '\n', X.shape
, '\n###############################################################')

6
scripts/ml/FS.py → scripts/ml/ml_functions/FS.py
View file

@ -4,7 +4,8 @@
Created on Mon May 23 23:25:26 2022
@author: tanu
"""
"""
#%%
import os, sys
import pandas as pd
import numpy as np
@ -389,4 +390,5 @@ def fsgs_rfecv(input_df
, '\nOutput dict size:', len(output_modelD))
return(output_modelD)
else:
sys.exit('\nFAIL:numbers mismatch output dict length not as expected. Please check')
sys.exit('\nFAIL:numbers mismatch output dict length not as expected. Please check')

1
scripts/ml/functions/GetMLData.py → scripts/ml/ml_functions/GetMLData.py
View file

@ -37,6 +37,7 @@ from sklearn.pipeline import Pipeline, make_pipeline
import argparse
import re
def getmldata(gene, drug
, data_combined_model = False
, use_or = False

533
scripts/ml/ml_functions/MultClfs.py Executable file
View file

@ -0,0 +1,533 @@
#!/usr/bin/env python3
# -*- coding: utf-8 -*-
"""
Created on Fri Mar 4 15:25:33 2022
@author: tanu
"""
#%%
import os, sys
import pandas as pd
import numpy as np
import pprint as pp
from copy import deepcopy
from sklearn import linear_model
from sklearn import datasets
from collections import Counter
from sklearn.linear_model import LogisticRegression, LogisticRegressionCV
from sklearn.linear_model import RidgeClassifier, RidgeClassifierCV, SGDClassifier, PassiveAggressiveClassifier
from sklearn.naive_bayes import BernoulliNB
from sklearn.neighbors import KNeighborsClassifier
from sklearn.svm import SVC
from sklearn.tree import DecisionTreeClassifier, ExtraTreeClassifier
from sklearn.ensemble import RandomForestClassifier, ExtraTreesClassifier, AdaBoostClassifier, GradientBoostingClassifier, BaggingClassifier
from sklearn.naive_bayes import GaussianNB
from sklearn.gaussian_process import GaussianProcessClassifier, kernels
from sklearn.gaussian_process.kernels import RBF, DotProduct, Matern, RationalQuadratic, WhiteKernel
from sklearn.discriminant_analysis import LinearDiscriminantAnalysis, QuadraticDiscriminantAnalysis
from sklearn.neural_network import MLPClassifier
from sklearn.svm import SVC
from xgboost import XGBClassifier
from sklearn.naive_bayes import MultinomialNB
from sklearn.preprocessing import StandardScaler, MinMaxScaler, OneHotEncoder
from sklearn.compose import ColumnTransformer
from sklearn.compose import make_column_transformer
from sklearn.metrics import make_scorer, confusion_matrix, accuracy_score, balanced_accuracy_score, precision_score, average_precision_score, recall_score
from sklearn.metrics import roc_auc_score, roc_curve, f1_score, matthews_corrcoef, jaccard_score, classification_report
# added
from sklearn.model_selection import train_test_split, cross_validate, cross_val_score, LeaveOneOut, KFold, RepeatedKFold, cross_val_predict
from sklearn.model_selection import train_test_split, cross_validate, cross_val_score
from sklearn.model_selection import StratifiedKFold,RepeatedStratifiedKFold, RepeatedKFold
from sklearn.pipeline import Pipeline, make_pipeline
from sklearn.feature_selection import RFE, RFECV
import itertools
import seaborn as sns
import matplotlib.pyplot as plt
from statistics import mean, stdev, median, mode
from imblearn.over_sampling import RandomOverSampler
from imblearn.under_sampling import RandomUnderSampler
from imblearn.over_sampling import SMOTE
from sklearn.datasets import make_classification
from imblearn.combine import SMOTEENN
from imblearn.combine import SMOTETomek
from imblearn.over_sampling import SMOTENC
from imblearn.under_sampling import EditedNearestNeighbours
from imblearn.under_sampling import RepeatedEditedNearestNeighbours
from sklearn.model_selection import GridSearchCV
from sklearn.base import BaseEstimator
from sklearn.impute import KNNImputer as KNN
import json
import argparse
import re
#%% GLOBALS
rs = {'random_state': 42}
njobs = {'n_jobs': 10}
scoring_fn = ({ 'mcc' : make_scorer(matthews_corrcoef)
, 'fscore' : make_scorer(f1_score)
, 'precision' : make_scorer(precision_score)
, 'recall' : make_scorer(recall_score)
, 'accuracy' : make_scorer(accuracy_score)
, 'roc_auc' : make_scorer(roc_auc_score)
, 'jcc' : make_scorer(jaccard_score)
})
skf_cv = StratifiedKFold(n_splits = 10
#, shuffle = False, random_state= None)
, shuffle = True,**rs)
rskf_cv = RepeatedStratifiedKFold(n_splits = 10
, n_repeats = 3
, **rs)
mcc_score_fn = {'mcc': make_scorer(matthews_corrcoef)}
jacc_score_fn = {'jcc': make_scorer(jaccard_score)}
###############################################################################
score_type_ordermapD = { 'mcc' : 1
, 'fscore' : 2
, 'jcc' : 3
, 'precision' : 4
, 'recall' : 5
, 'accuracy' : 6
, 'roc_auc' : 7
, 'TN' : 8
, 'FP' : 9
, 'FN' : 10
, 'TP' : 11
, 'trainingY_neg': 12
, 'trainingY_pos': 13
, 'blindY_neg' : 14
, 'blindY_pos' : 15
, 'fit_time' : 16
, 'score_time' : 17
}
scoreCV_mapD = {'test_mcc' : 'MCC'
, 'test_fscore' : 'F1'
, 'test_precision' : 'Precision'
, 'test_recall' : 'Recall'
, 'test_accuracy' : 'Accuracy'
, 'test_roc_auc' : 'ROC_AUC'
, 'test_jcc' : 'JCC'
}
scoreBT_mapD = {'bts_mcc' : 'MCC'
, 'bts_fscore' : 'F1'
, 'bts_precision' : 'Precision'
, 'bts_recall' : 'Recall'
, 'bts_accuracy' : 'Accuracy'
, 'bts_roc_auc' : 'ROC_AUC'
, 'bts_jcc' : 'JCC'
}
#%%############################################################################
############################
# MultModelsCl()
# Run Multiple Classifiers
############################
# Multiple Classification - Model Pipeline
def MultModelsCl(input_df, target, skf_cv
, blind_test_df
, blind_test_target
, tts_split_type
, run_blind_test = True
, resampling_type = 'none' # default
, add_cm = True # adds confusion matrix based on cross_val_predict
, add_yn = True # adds target var class numbers
, var_type = ['numerical', 'categorical','mixed']
, return_formatted_output = True):
'''
@ param input_df: input features
@ type: df with input features WITHOUT the target variable
@param target: target (or output) feature
@type: df or np.array or Series
@param skv_cv: stratifiedK fold int or object to allow shuffle and random state to pass
@type: int or StratifiedKfold()
@var_type: numerical, categorical and mixed to determine what col_transform to apply (MinMaxScalar and/or one-ho t encoder)
@type: list
returns
Dict containing multiple classification scores for each model and mean of each Stratified Kfold including training
'''
#======================================================
# Determine categorical and numerical features
#======================================================
numerical_ix = input_df.select_dtypes(include=['int64', 'float64']).columns
numerical_ix
categorical_ix = input_df.select_dtypes(include=['object', 'bool']).columns
categorical_ix
#======================================================
# Determine preprocessing steps ~ var_type
#======================================================
if var_type == 'numerical':
t = [('num', MinMaxScaler(), numerical_ix)]
if var_type == 'categorical':
t = [('cat', OneHotEncoder(), categorical_ix)]
if var_type == 'mixed':
t = [('num', MinMaxScaler(), numerical_ix)
, ('cat', OneHotEncoder(), categorical_ix) ]
col_transform = ColumnTransformer(transformers = t
, remainder='passthrough')
#======================================================
# Specify multiple Classification Models
#======================================================
models = [('AdaBoost Classifier' , AdaBoostClassifier(**rs) )
, ('Bagging Classifier' , BaggingClassifier(**rs, **njobs, bootstrap = True, oob_score = True) )
, ('Decision Tree' , DecisionTreeClassifier(**rs) )
, ('Extra Tree' , ExtraTreeClassifier(**rs) )
, ('Extra Trees' , ExtraTreesClassifier(**rs) )
, ('Gradient Boosting' , GradientBoostingClassifier(**rs) )
, ('Gaussian NB' , GaussianNB() )
, ('Gaussian Process' , GaussianProcessClassifier(**rs) )
, ('K-Nearest Neighbors' , KNeighborsClassifier() )
, ('LDA' , LinearDiscriminantAnalysis() )
, ('Logistic Regression' , LogisticRegression(**rs) )
, ('Logistic RegressionCV' , LogisticRegressionCV(cv = 3, **rs))
, ('MLP' , MLPClassifier(max_iter = 500, **rs) )
, ('Multinomial' , MultinomialNB() )
, ('Naive Bayes' , BernoulliNB() )
, ('Passive Aggresive' , PassiveAggressiveClassifier(**rs, **njobs) )
, ('QDA' , QuadraticDiscriminantAnalysis() )
, ('Random Forest' , RandomForestClassifier(**rs, n_estimators = 1000 ) )
, ('Random Forest2' , RandomForestClassifier(min_samples_leaf = 5
, n_estimators = 1000
, bootstrap = True
, oob_score = True
, **njobs
, **rs
, max_features = 'auto') )
, ('Ridge Classifier' , RidgeClassifier(**rs) )
, ('Ridge ClassifierCV' , RidgeClassifierCV(cv = 3) )
, ('SVC' , SVC(**rs) )
, ('Stochastic GDescent' , SGDClassifier(**rs, **njobs) )
, ('XGBoost' , XGBClassifier(**rs, verbosity = 0, use_label_encoder =False) )
]
mm_skf_scoresD = {}
print('\n==============================================================\n'
, '\nRunning several classification models (n):', len(models)
,'\nList of models:')
for m in models:
print(m)
print('\n================================================================\n')
index = 1
for model_name, model_fn in models:
print('\nRunning classifier:', index
, '\nModel_name:' , model_name
, '\nModel func:' , model_fn)
index = index+1
model_pipeline = Pipeline([
('prep' , col_transform)
, ('model' , model_fn)])
print('\nRunning model pipeline:', model_pipeline)
skf_cv_modD = cross_validate(model_pipeline
, input_df
, target
, cv = skf_cv
, scoring = scoring_fn
, return_train_score = True)
#==============================
# Extract mean values for CV
#==============================
mm_skf_scoresD[model_name] = {}
for key, value in skf_cv_modD.items():
print('\nkey:', key, '\nvalue:', value)
print('\nmean value:', np.mean(value))
mm_skf_scoresD[model_name][key] = round(np.mean(value),2)
# ADD more info: meta data related to input df
mm_skf_scoresD[model_name]['resampling'] = resampling_type
mm_skf_scoresD[model_name]['n_training_size'] = len(input_df)
mm_skf_scoresD[model_name]['n_trainingY_ratio'] = round(Counter(target)[0]/Counter(target)[1], 2)
mm_skf_scoresD[model_name]['n_features'] = len(input_df.columns)
mm_skf_scoresD[model_name]['tts_split'] = tts_split_type
#######################################################################
#======================================================
# Option: Add confusion matrix from cross_val_predict
# Understand and USE with caution
#======================================================
if add_cm:
cmD = {}
# Calculate cm
y_pred = cross_val_predict(model_pipeline, input_df, target, cv = skf_cv, **njobs)
#_tn, _fp, _fn, _tp = confusion_matrix(y_pred, y).ravel() # internally
tn, fp, fn, tp = confusion_matrix(y_pred, target).ravel()
# Build cm dict
cmD = {'TN' : tn
, 'FP': fp
, 'FN': fn
, 'TP': tp}
# Update cv dict cmD
mm_skf_scoresD[model_name].update(cmD)
#=============================================
# Option: Add targety numbers for data
#=============================================
if add_yn:
tnD = {}
# Build tn numbers dict
tnD = {'n_trainingY_neg' : Counter(target)[0]
, 'n_trainingY_pos' : Counter(target)[1] }
# Update cv dict with cmD and tnD
mm_skf_scoresD[model_name].update(tnD)
#%%
#=========================
# Option: Blind test (bts)
#=========================
if run_blind_test:
btD = {}
# Build bts numbers dict
btD = {'n_blindY_neg' : Counter(blind_test_target)[0]
, 'n_blindY_pos' : Counter(blind_test_target)[1]
, 'n_testY_ratio' : round(Counter(blind_test_target)[0]/Counter(blind_test_target)[1], 2)
, 'n_test_size' : len(blind_test_df) }
# Update cmD+tnD dicts with btD
mm_skf_scoresD[model_name].update(btD)
#--------------------------------------------------------
# Build the final results with all scores for the model
#--------------------------------------------------------
#bts_predict = gscv_fs.predict(blind_test_df)
model_pipeline.fit(input_df, target)
bts_predict = model_pipeline.predict(blind_test_df)
bts_mcc_score = round(matthews_corrcoef(blind_test_target, bts_predict),2)
print('\nMCC on Blind test:' , bts_mcc_score)
print('\nAccuracy on Blind test:', round(accuracy_score(blind_test_target, bts_predict),2))
mm_skf_scoresD[model_name]['bts_mcc'] = bts_mcc_score
mm_skf_scoresD[model_name]['bts_fscore'] = round(f1_score(blind_test_target, bts_predict),2)
mm_skf_scoresD[model_name]['bts_precision'] = round(precision_score(blind_test_target, bts_predict),2)
mm_skf_scoresD[model_name]['bts_recall'] = round(recall_score(blind_test_target, bts_predict),2)
mm_skf_scoresD[model_name]['bts_accuracy'] = round(accuracy_score(blind_test_target, bts_predict),2)
mm_skf_scoresD[model_name]['bts_roc_auc'] = round(roc_auc_score(blind_test_target, bts_predict),2)
mm_skf_scoresD[model_name]['bts_jcc'] = round(jaccard_score(blind_test_target, bts_predict),2)
#mm_skf_scoresD[model_name]['diff_mcc'] = train_test_diff_MCC
#%%
# ADD more info: meta data related to input and blind and resampling
# target numbers: training
yc1 = Counter(target)
yc1_ratio = yc1[0]/yc1[1]
# target numbers: test
yc2 = Counter(blind_test_target)
yc2_ratio = yc2[0]/yc2[1]
mm_skf_scoresD[model_name]['resampling'] = resampling_type
mm_skf_scoresD[model_name]['n_training_size'] = len(input_df)
mm_skf_scoresD[model_name]['n_trainingY_ratio'] = round(yc1_ratio, 2)
mm_skf_scoresD[model_name]['n_test_size'] = len(blind_test_df)
mm_skf_scoresD[model_name]['n_testY_ratio'] = round(yc2_ratio,2)
mm_skf_scoresD[model_name]['n_features'] = len(input_df.columns)
mm_skf_scoresD[model_name]['tts_split'] = tts_split_type
#return(mm_skf_scoresD)
#============================
# Process the dict to have WF
#============================
if return_formatted_output:
CV_BT_metaDF = ProcessMultModelsCl(mm_skf_scoresD)
return(CV_BT_metaDF)
else:
return(mm_skf_scoresD)
#%% Process output function ###################################################
############################
# ProcessMultModelsCl()
############################
#Processes the dict from above if use_formatted_output = True
def ProcessMultModelsCl(inputD = {}, blind_test_data = True):
scoresDF = pd.DataFrame(inputD)
#------------------------
# Extracting split_name
#-----------------------
tts_split_nameL = []
for k,v in inputD.items():
tts_split_nameL = tts_split_nameL + [v['tts_split']]
if len(set(tts_split_nameL)) == 1:
tts_split_name = str(list(set(tts_split_nameL))[0])
print('\nExtracting tts_split_name:', tts_split_name)
#----------------------
# WF: CV results
#----------------------
scoresDFT = scoresDF.T
scoresDF_CV = scoresDFT.filter(regex='^test_.*$', axis = 1); scoresDF_CV.columns
# map colnames for consistency to allow concatenting
scoresDF_CV.columns = scoresDF_CV.columns.map(scoreCV_mapD); scoresDF_CV.columns
scoresDF_CV['source_data'] = 'CV'
#----------------------
# WF: Meta data
#----------------------
metaDF = scoresDFT.filter(regex='^(?!test_.*$|bts_.*$|train_.*$).*'); metaDF.columns
print('\nTotal cols in each df:'
, '\nCV df:', len(scoresDF_CV.columns)
, '\nmetaDF:', len(metaDF.columns))
#-------------------------------------
# Combine WF: CV + Metadata
#-------------------------------------
combDF = pd.merge(scoresDF_CV, metaDF, left_index = True, right_index = True)
print('\nAdding column: Model_name')
combDF['Model_name'] = combDF.index
#----------------------
# WF: BTS results
#----------------------
if blind_test_data:
scoresDF_BT = scoresDFT.filter(regex='^bts_.*$', axis = 1); scoresDF_BT.columns
# map colnames for consistency to allow concatenting
scoresDF_BT.columns = scoresDF_BT.columns.map(scoreBT_mapD); scoresDF_BT.columns
scoresDF_BT['source_data'] = 'BT'
print('\nTotal cols in bts df:'
, '\nBT_df:', len(scoresDF_BT.columns))
if len(scoresDF_CV.columns) == len(scoresDF_BT.columns):
print('\nFirst proceeding to rowbind CV and BT dfs:')
expected_ncols_out = len(scoresDF_BT.columns) + len(metaDF.columns)
print('\nFinal output should have:', expected_ncols_out, 'columns' )
#-----------------
# Combine WF
#-----------------
dfs_combine_wf = [scoresDF_CV, scoresDF_BT]
print('\nCombinig', len(dfs_combine_wf), 'using pd.concat by row ~ rowbind'
, '\nChecking Dims of df to combine:'
, '\nDim of CV:', scoresDF_CV.shape
, '\nDim of BT:', scoresDF_BT.shape)
#print(scoresDF_CV)
#print(scoresDF_BT)
dfs_nrows_wf = []
for df in dfs_combine_wf:
dfs_nrows_wf = dfs_nrows_wf + [len(df)]
dfs_nrows_wf = max(dfs_nrows_wf)
dfs_ncols_wf = []
for df in dfs_combine_wf:
dfs_ncols_wf = dfs_ncols_wf + [len(df.columns)]
dfs_ncols_wf = max(dfs_ncols_wf)
print(dfs_ncols_wf)
expected_nrows_wf = len(dfs_combine_wf) * dfs_nrows_wf
expected_ncols_wf = dfs_ncols_wf
common_cols_wf = list(set.intersection(*(set(df.columns) for df in dfs_combine_wf)))
print('\nNumber of Common columns:', dfs_ncols_wf
, '\nThese are:', common_cols_wf)
if len(common_cols_wf) == dfs_ncols_wf :
combined_baseline_wf = pd.concat([df[common_cols_wf] for df in dfs_combine_wf], ignore_index=False)
print('\nConcatenating dfs with different resampling methods [WF]:'
, '\nSplit type:', tts_split_name
, '\nNo. of dfs combining:', len(dfs_combine_wf))
#print('\n================================================^^^^^^^^^^^^')
if len(combined_baseline_wf) == expected_nrows_wf and len(combined_baseline_wf.columns) == expected_ncols_wf:
#print('\n================================================^^^^^^^^^^^^')
print('\nPASS:', len(dfs_combine_wf), 'dfs successfully combined'
, '\nnrows in combined_df_wf:', len(combined_baseline_wf)
, '\nncols in combined_df_wf:', len(combined_baseline_wf.columns))
else:
print('\nFAIL: concatenating failed'
, '\nExpected nrows:', expected_nrows_wf
, '\nGot:', len(combined_baseline_wf)
, '\nExpected ncols:', expected_ncols_wf
, '\nGot:', len(combined_baseline_wf.columns))
sys.exit('\nFIRST IF FAILS')
##
c1L = list(set(combined_baseline_wf.index))
c2L = list(metaDF.index)
#if set(c1L) == set(c2L):
if set(c1L) == set(c2L) and all(x in c2L for x in c1L) and all(x in c1L for x in c2L):
print('\nPASS: proceeding to merge metadata with CV and BT dfs')
combDF = pd.merge(combined_baseline_wf, metaDF, left_index = True, right_index = True)
print('\nAdding column: Model_name')
combDF['Model_name'] = combDF.index
else:
sys.exit('\nFAIL: Could not merge metadata with CV and BT dfs')
else:
print('\nConcatenting dfs not possible [WF],check numbers ')
#-------------------------------------
# Combine WF+Metadata: Final output
#-------------------------------------
# if len(combDF.columns) == expected_ncols_out:
# print('\nPASS: Combined df has expected ncols')
# else:
# sys.exit('\nFAIL: Length mismatch for combined_df')
# print('\nAdding column: Model_name')
# combDF['Model_name'] = combDF.index
print('\n========================================================='
, '\nSUCCESS: Ran multiple classifiers'
, '\n=======================================================')
#resampling_methods_wf = combined_baseline_wf[['resampling']]
#resampling_methods_wf = resampling_methods_wf.drop_duplicates()
#, '\n', resampling_methods_wf)
return combDF
###############################################################################

122
scripts/ml/combined_model/untitled0.py → scripts/ml/ml_functions/SplitTTS.py
View file

@ -39,16 +39,21 @@ from sklearn.pipeline import Pipeline, make_pipeline
import argparse
import re
homedir = os.path.expanduser("~")
#%% Globals
#%% GLOBALS
rs = {'random_state': 42}
njobs = {'n_jobs': 10}
#%% Define split_tts function #################################################
def split_tts(ml_input_data
, data_type = ['actual', 'complete', 'reverse']
, data_type = ['actual', 'complete']
, split_type = ['70_30', '80_20', 'sl']
, oversampling = True
, dst_colname = 'dst'# determine how to subset the actual vs reverse data
, target_colname = 'dst_mode'):
, target_colname = 'dst_mode'
, include_gene_name = True
, k_smote = 5):
outDict = {}
print('\nInput params:'
, '\nDim of input df:' , ml_input_data.shape
@ -60,6 +65,11 @@ def split_tts(ml_input_data
print('\noversampling enabled')
else:
print('\nNot generating oversampled or undersampled data')
if include_gene_name:
cols_to_dropL = []
else:
cols_to_dropL = ['gene_name']
#====================================
# evaluating use_data_type
@ -68,21 +78,26 @@ def split_tts(ml_input_data
ml_data = ml_input_data[ml_input_data[dst_colname].notna()]
if data_type == 'complete':
ml_data = ml_input_data.copy()
if data_type == 'reverse':
ml_data = ml_input_data[ml_input_data[dst_colname].isna()]
#if_data_type == none
#====================================
# separate features and target
#====================================
x_features = ml_data.drop([target_colname, dst_colname], axis = 1)
y_target = ml_data[target_colname]
cols_to_dropL = cols_to_dropL + [target_colname, dst_colname]
x_features = ml_data.drop(cols_to_dropL, axis = 1)
y_target = ml_data[target_colname]
# sanity check
if not 'dst_mode' in x_features.columns:
print('\nPASS: x_features has no target variable')
check1 = x_features[[i for i in cols_to_dropL if i in x_features.columns]]
#if not 'dst_mode' in x_features.columns:
if check1.empty:
print('\nPASS: x_features has no target variable and no dst column'
, '\nDropped cols:', len(cols_to_dropL)
, '\nThese were:', target_colname,'and', dst_colname)
x_ncols = len(x_features.columns)
print('\nNo. of columns for x_features:', x_ncols)
print('\nNo. of cols in input df:', len(ml_input_data.columns)
, '\nNo.of cols dropped:', len(cols_to_dropL)
, '\nNo. of columns for x_features:', x_ncols)
else:
sys.exit('\nFAIL: x_features has target variable included. FIX it and rerun!')
@ -129,7 +144,12 @@ def split_tts(ml_input_data
, '\n\nTotal no. of input features:' , len(X.columns)
, '\n--------No. of numerical features:' , len(numerical_cols)
, '\n--------No. of categorical features:', len(categorical_cols)
, '\n==========================='
, '\n Resampling: NONE'
, '\nBaseline'
, '\n==========================='
, '\n\nTotal data size:', len(X) + len(X_bts)
, '\n\nTrain data size:', X.shape
@ -140,11 +160,15 @@ def split_tts(ml_input_data
, '\n\ny_train ratio:',yc1_ratio
, '\ny_test ratio:', yc2_ratio
, '\n-------------------------------------------------------------'
)
, '\n-------------------------------------------------------------')
outDict.update({'X' : X
, 'X_bts' : X_bts
, 'y' : y
, 'y_bts' : y_bts
} )
if oversampling:
#######################################################################
# RESAMPLING
#######################################################################
@ -194,28 +218,70 @@ def split_tts(ml_input_data
categorical_colind = X.columns.get_indexer(list(categorical_ix))
categorical_colind
k_sm = 5 # default
#k_sm = 5 # default
k_sm = k_smote
sm_nc = SMOTENC(categorical_features=categorical_colind, k_neighbors = k_sm, **rs, **njobs)
X_smnc, y_smnc = sm_nc.fit_resample(X, y)
print('\nSMOTE_NC OverSampling\n', Counter(y_smnc))
print(X_smnc.shape)
print('\nGenerated resampled data as below:'
, '\n==========================='
, '\nRandom oversampling:'
, '\n==========================='
print('\nGenerated Resampled data as below:'
, '\n================================='
, '\nResampling: Random oversampling'
, '\n================================'
, '\n\nTrain data size:', X_ros.shape
, '\ny_train numbers:', y_ros
, '\n\ny_train ratio:', Counter(y_ros)[0]/Counter(y_ros)[0]
, '\ny_train numbers:', len(y_ros)
, '\n\ny_train ratio:', Counter(y_ros)[0]/Counter(y_ros)[1]
, '\ny_test ratio:' , yc2_ratio
##################################################################
, '\n================================'
, '\nResampling: Random underampling'
, '\n================================'
, '\n-------------------------------------------------------------'
)
, '\n\nTrain data size:', X_rus.shape
, '\ny_train numbers:', len(y_rus)
, '\n\ny_train ratio:', Counter(y_rus)[0]/Counter(y_rus)[1]
, '\ny_test ratio:' , yc2_ratio
##################################################################
, '\n================================'
, '\nResampling:Combined (over+under)'
, '\n================================'
, '\n\nTrain data size:', X_rouC.shape
, '\ny_train numbers:', len(y_rouC)
, '\n\ny_train ratio:', Counter(y_rouC)[0]/Counter(y_rouC)[1]
, '\ny_test ratio:' , yc2_ratio
##################################################################
, '\n=============================='
, '\nResampling: Smote NC'
, '\n=============================='
, '\n\nTrain data size:', X_smnc.shape
, '\ny_train numbers:', len(y_smnc)
, '\n\ny_train ratio:', Counter(y_smnc)[0]/Counter(y_smnc)[1]
, '\ny_test ratio:' , yc2_ratio
##################################################################
, '\n-------------------------------------------------------------')
outDict.update({'X_ros' : X_ros
, 'y_ros' : y_ros
, 'X_rus' : X_rus
, 'y_rus' : y_rus
, 'X_rouC': X_rouC
, 'y_rouC': y_rouC
, 'X_smnc': X_smnc
, 'y_smnc': y_smnc})
return(outDict)
# globals().update(locals()) # TROLOLOLOLOLOLS
#return()
else:
return(outDict)

141
scripts/ml/run_7030.py
View file

@ -1,141 +0,0 @@
#!/usr/bin/env python3
# -*- coding: utf-8 -*-
"""
Created on Mon Jun 20 13:05:23 2022
@author: tanu
"""
#%%Imports ####################################################################
import re
import argparse
import os, sys
# gene = 'pncA'
# drug = 'pyrazinamide'
#total_mtblineage_uc = 8
###############################################################################
#%% command line args: case sensitive
arg_parser = argparse.ArgumentParser()
arg_parser.add_argument('-d', '--drug', help = 'drug name', default = '')
arg_parser.add_argument('-g', '--gene', help = 'gene name', default = '')
args = arg_parser.parse_args()
drug = args.drug
gene = args.gene
###############################################################################
homedir = os.path.expanduser("~")
sys.path.append(homedir + '/git/LSHTM_analysis/scripts/ml')
###############################################################################
#==================
# Import data
#==================
from ml_data_7030 import *
setvars(gene,drug)
from ml_data_7030 import *
# from YC run_all_ML: run locally
#from UQ_yc_RunAllClfs import run_all_ML
#====================
# Import ML functions
#====================
from MultClfs import *
#==================
# other vars
#==================
tts_split_7030 = '70_30'
OutFile_suffix = '7030'
#==================
# Specify outdir
#==================
outdir_ml = outdir + 'ml/tts_7030/'
print('\nOutput directory:', outdir_ml)
#outFile_wf = outdir_ml + gene.lower() + '_baselineC_' + OutFile_suffix + '.csv'
outFile_wf = outdir_ml + gene.lower() + '_baselineC_noOR' + OutFile_suffix + '.csv'
#%% Running models ############################################################
print('\n#####################################################################\n'
, '\nStarting--> Running ML analysis: Baseline modes (No FS)'
, '\nGene name:', gene
, '\nDrug name:', drug
, '\n#####################################################################\n')
paramD = {
'baseline_paramD': { 'input_df' : X
, 'target' : y
, 'var_type' : 'mixed'
, 'resampling_type': 'none'}
, 'smnc_paramD': { 'input_df' : X_smnc
, 'target' : y_smnc
, 'var_type' : 'mixed'
, 'resampling_type' : 'smnc'}
, 'ros_paramD': { 'input_df' : X_ros
, 'target' : y_ros
, 'var_type' : 'mixed'
, 'resampling_type' : 'ros'}
, 'rus_paramD' : { 'input_df' : X_rus
, 'target' : y_rus
, 'var_type' : 'mixed'
, 'resampling_type' : 'rus'}
, 'rouC_paramD' : { 'input_df' : X_rouC
, 'target' : y_rouC
, 'var_type' : 'mixed'
, 'resampling_type' : 'rouC'}
}
##==============================================================================
## Dict with no CV BT formatted df
## mmD = {}
## for k, v in paramD.items():
## # print(mmD[k])
## scores_7030D = MultModelsCl(**paramD[k]
## , tts_split_type = tts_split_7030
## , skf_cv = skf_cv
## , blind_test_df = X_bts
## , blind_test_target = y_bts
## , add_cm = True
## , add_yn = True
## , return_formatted_output = False)
## mmD[k] = scores_7030D
##==============================================================================
## Initial run to get the dict of dicts for each sampling type containing CV, BT and metadata DFs
mmDD = {}
for k, v in paramD.items():
scores_7030D = MultModelsCl(**paramD[k]
, tts_split_type = tts_split_7030
, skf_cv = skf_cv
, blind_test_df = X_bts
, blind_test_target = y_bts
, add_cm = True
, add_yn = True
, return_formatted_output = True)
mmDD[k] = scores_7030D
# Extracting the dfs from within the dict and concatenating to output as one df
for k, v in mmDD.items():
out_wf_7030 = pd.concat(mmDD, ignore_index = True)
out_wf_7030f = out_wf_7030.sort_values(by = ['resampling', 'source_data', 'MCC'], ascending = [True, True, False], inplace = False)
print('\n######################################################################'
, '\nEnd--> Successfully generated output DF for Multiple classifiers (baseline models)'
, '\nGene:', gene.lower()
, '\nDrug:', drug
, '\noutput file:', outFile_wf
, '\nDim of output:', out_wf_7030f.shape
, '\n######################################################################')
###############################################################################
#====================
# Write output file
#====================
out_wf_7030f.to_csv(outFile_wf, index = False)
print('\nFile successfully written:', outFile_wf)
###############################################################################

126
scripts/ml/run_7030_LOOP.py
View file

@ -1,126 +0,0 @@
#!/usr/bin/env python3
# -*- coding: utf-8 -*-
"""
Created on Mon Jun 20 13:05:23 2022
@author: tanu
"""
#%%Imports ####################################################################
import re
import argparse
import os, sys
import collections
# gene = 'pncA'
# drug = 'pyrazinamide'
#total_mtblineage_uc = 8
###############################################################################
#%% command line args: case sensitive
# arg_parser = argparse.ArgumentParser()
# arg_parser.add_argument('-d', '--drug', help = 'drug name', default = '')
# arg_parser.add_argument('-g', '--gene', help = 'gene name', default = '')
# args = arg_parser.parse_args()
# drug = args.drug
# gene = args.gene
###############################################################################
homedir = os.path.expanduser("~")
sys.path.append(homedir + '/git/LSHTM_analysis/scripts/ml')
###############################################################################
#==================
# Import data
#==================
from ml_data_7030 import *
setvars(gene,drug)
from ml_data_7030 import *
# from YC run_all_ML: run locally
#from UQ_yc_RunAllClfs import run_all_ML
#====================
# Import ML functions
#====================
from MultClfs import *
#==================
# other vars
#==================
tts_split_7030 = '70_30'
OutFile_suffix = '7030'
#==================
# Specify outdir
#==================
outdir_ml = outdir + 'ml/tts_7030/'
print('\nOutput directory:', outdir_ml)
outFile_wf = outdir_ml + gene.lower() + '_baselineC_' + OutFile_suffix + '.csv'
#outFile_lf = outdir_ml + gene.lower() + '_baselineC_ext_' + OutFile_suffix + '.csv'
#%% Running models ############################################################
print('\n#####################################################################\n'
, '\nStarting--> Running ML analysis: Baseline modes (No FS)'
, '\nGene name:', gene
, '\nDrug name:', drug
, '\n#####################################################################\n')
paramD = {
'baseline_paramD': { 'input_df' : X
, 'target' : y
, 'var_type' : 'mixed'
, 'resampling_type': 'none'}
, 'smnc_paramD': { 'input_df' : X_smnc
, 'target' : y_smnc
, 'var_type' : 'mixed'
, 'resampling_type' : 'smnc'}
, 'ros_paramD': { 'input_df' : X_ros
, 'target' : y_ros
, 'var_type' : 'mixed'
, 'resampling_type' : 'ros'}
, 'rus_paramD' : { 'input_df' : X_rus
, 'target' : y_rus
, 'var_type' : 'mixed'
, 'resampling_type' : 'rus'}
, 'rouC_paramD' : { 'input_df' : X_rouC
, 'target' : y_rouC
, 'var_type' : 'mixed'
, 'resampling_type' : 'rouC'}
}
# Initial run to get the dict containing CV, BT and metadata DFs
mmD = {}
for k, v in paramD.items():
# print(fooD[k])
scores_7030D = MultModelsCl(**paramD[k]
, tts_split_type = tts_split_7030
, skf_cv = skf_cv
, blind_test_df = X_bts
, blind_test_target = y_bts
, add_cm = True
, add_yn = True
, return_formatted_output = True)
mmD[k] = scores_7030D
for k, v in mmD.items():
out_wf_7030 = pd.concat(mmD, ignore_index = True)
print('\n######################################################################'
, '\nEnd--> Successfully generated output DF for Multiple classifiers (baseline models)'
, '\nGene:', gene.lower()
, '\nDrug:', drug
, '\noutput file:', outFile_wf
, '\nDim of output:', out_wf_7030.shape
, '\n######################################################################')
###############################################################################
#====================
# Write output file
#====================
out_wf_7030.to_csv(outFile_wf, index = False)
print('\nFile successfully written:', outFile_wf)
###############################################################################

141
scripts/ml/run_8020.py
View file

@ -1,141 +0,0 @@
#!/usr/bin/env python3
# -*- coding: utf-8 -*-
"""
Created on Mon Jun 20 13:05:23 2022
@author: tanu
"""
#%%Imports ####################################################################
import re
import argparse
import os, sys
# gene = 'pncA'
# drug = 'pyrazinamide'
#total_mtblineage_uc = 8
###############################################################################
#%% command line args: case sensitive
arg_parser = argparse.ArgumentParser()
arg_parser.add_argument('-d', '--drug', help = 'drug name', default = '')
arg_parser.add_argument('-g', '--gene', help = 'gene name', default = '')
args = arg_parser.parse_args()
drug = args.drug
gene = args.gene
###############################################################################
homedir = os.path.expanduser("~")
sys.path.append(homedir + '/git/LSHTM_analysis/scripts/ml')
###############################################################################
#==================
# Import data
#==================
from ml_data_8020 import *
setvars(gene,drug)
from ml_data_8020 import *
# from YC run_all_ML: run locally
#from UQ_yc_RunAllClfs import run_all_ML
#====================
# Import ML functions
#====================
from MultClfs import *
#==================
# other vars
#==================
tts_split_8020 = '80_20'
OutFile_suffix = '8020'
#==================
# Specify outdir
#==================
outdir_ml = outdir + 'ml/tts_8020/'
print('\nOutput directory:', outdir_ml)
#outFile_wf = outdir_ml + gene.lower() + '_baselineC_' + OutFile_suffix + '.csv'
outFile_wf = outdir_ml + gene.lower() + '_baselineC_noOR' + OutFile_suffix + '.csv'
#%% Running models ############################################################
print('\n#####################################################################\n'
, '\nStarting--> Running ML analysis: Baseline modes (No FS)'
, '\nGene name:', gene
, '\nDrug name:', drug
, '\n#####################################################################\n')
paramD = {
'baseline_paramD': { 'input_df' : X
, 'target' : y
, 'var_type' : 'mixed'
, 'resampling_type': 'none'}
, 'smnc_paramD': { 'input_df' : X_smnc
, 'target' : y_smnc
, 'var_type' : 'mixed'
, 'resampling_type' : 'smnc'}
, 'ros_paramD': { 'input_df' : X_ros
, 'target' : y_ros
, 'var_type' : 'mixed'
, 'resampling_type' : 'ros'}
, 'rus_paramD' : { 'input_df' : X_rus
, 'target' : y_rus
, 'var_type' : 'mixed'
, 'resampling_type' : 'rus'}
, 'rouC_paramD' : { 'input_df' : X_rouC
, 'target' : y_rouC
, 'var_type' : 'mixed'
, 'resampling_type' : 'rouC'}
}
##==============================================================================
## Dict with no CV BT formatted df
## mmD = {}
## for k, v in paramD.items():
## # print(mmD[k])
## scores_8020D = MultModelsCl(**paramD[k]
## , tts_split_type = tts_split_8020
## , skf_cv = skf_cv
## , blind_test_df = X_bts
## , blind_test_target = y_bts
## , add_cm = True
## , add_yn = True
## , return_formatted_output = False)
## mmD[k] = scores_8020D
##==============================================================================
## Initial run to get the dict of dicts for each sampling type containing CV, BT and metadata DFs
mmDD = {}
for k, v in paramD.items():
scores_8020D = MultModelsCl(**paramD[k]
, tts_split_type = tts_split_8020
, skf_cv = skf_cv
, blind_test_df = X_bts
, blind_test_target = y_bts
, add_cm = True
, add_yn = True
, return_formatted_output = True)
mmDD[k] = scores_8020D
# Extracting the dfs from within the dict and concatenating to output as one df
for k, v in mmDD.items():
out_wf_8020 = pd.concat(mmDD, ignore_index = True)
out_wf_8020f = out_wf_8020.sort_values(by = ['resampling', 'source_data', 'MCC'], ascending = [True, True, False], inplace = False)
print('\n######################################################################'
, '\nEnd--> Successfully generated output DF for Multiple classifiers (baseline models)'
, '\nGene:', gene.lower()
, '\nDrug:', drug
, '\noutput file:', outFile_wf
, '\nDim of output:', out_wf_8020f.shape
, '\n######################################################################')
###############################################################################
#====================
# Write output file
#====================
out_wf_8020f.to_csv(outFile_wf, index = False)
print('\nFile successfully written:', outFile_wf)
###############################################################################

255
scripts/ml/run_FS.py
View file

@ -1,255 +0,0 @@
#!/usr/bin/env python3
# -*- coding: utf-8 -*-
"""
Created on Tue May 24 08:11:05 2022
@author: tanu
"""
#%%
import os, sys
import pandas as pd
import numpy as np
import pprint as pp
from copy import deepcopy
from sklearn import linear_model
from sklearn import datasets
from collections import Counter
from sklearn.linear_model import LogisticRegression, LogisticRegressionCV
from sklearn.linear_model import RidgeClassifier, RidgeClassifierCV, SGDClassifier, PassiveAggressiveClassifier
from sklearn.naive_bayes import BernoulliNB
from sklearn.neighbors import KNeighborsClassifier
from sklearn.svm import SVC
from sklearn.tree import DecisionTreeClassifier, ExtraTreeClassifier
from sklearn.ensemble import RandomForestClassifier, ExtraTreesClassifier, AdaBoostClassifier, GradientBoostingClassifier, BaggingClassifier
from sklearn.naive_bayes import GaussianNB
from sklearn.gaussian_process import GaussianProcessClassifier, kernels
from sklearn.gaussian_process.kernels import RBF, DotProduct, Matern, RationalQuadratic, WhiteKernel
from sklearn.discriminant_analysis import LinearDiscriminantAnalysis, QuadraticDiscriminantAnalysis
from sklearn.neural_network import MLPClassifier
from sklearn.svm import SVC
from xgboost import XGBClassifier
from sklearn.naive_bayes import MultinomialNB
from sklearn.preprocessing import StandardScaler, MinMaxScaler, OneHotEncoder
from sklearn.compose import ColumnTransformer
from sklearn.compose import make_column_transformer
from sklearn.metrics import make_scorer, confusion_matrix, accuracy_score, balanced_accuracy_score, precision_score, average_precision_score, recall_score
from sklearn.metrics import roc_auc_score, roc_curve, f1_score, matthews_corrcoef, jaccard_score, classification_report
# added
from sklearn.model_selection import train_test_split, cross_validate, cross_val_score, LeaveOneOut, KFold, RepeatedKFold, cross_val_predict
from sklearn.model_selection import train_test_split, cross_validate, cross_val_score
from sklearn.model_selection import StratifiedKFold,RepeatedStratifiedKFold, RepeatedKFold
from sklearn.pipeline import Pipeline, make_pipeline
from sklearn.feature_selection import RFE, RFECV
import itertools
import seaborn as sns
import matplotlib.pyplot as plt
from statistics import mean, stdev, median, mode
from imblearn.over_sampling import RandomOverSampler
from imblearn.under_sampling import RandomUnderSampler
from imblearn.over_sampling import SMOTE
from sklearn.datasets import make_classification
from imblearn.combine import SMOTEENN
from imblearn.combine import SMOTETomek
from imblearn.over_sampling import SMOTENC
from imblearn.under_sampling import EditedNearestNeighbours
from imblearn.under_sampling import RepeatedEditedNearestNeighbours
from sklearn.model_selection import GridSearchCV
from sklearn.base import BaseEstimator
from sklearn.impute import KNNImputer as KNN
import json
import argparse
import re
###############################################################################
#gene = 'pncA'
#drug = 'pyrazinamide'
#total_mtblineage_uc = 8
#%% command line args: case sensitive
arg_parser = argparse.ArgumentParser()
arg_parser.add_argument('-d', '--drug', help = 'drug name', default = '')
arg_parser.add_argument('-g', '--gene', help = 'gene name', default = '')
args = arg_parser.parse_args()
drug = args.drug
gene = args.gene
###############################################################################
#==================
# other vars
#==================
tts_split = '70_30'
OutFile_suffix = '7030_FS'
###############################################################################
#==================
# Import data
#==================
from ml_data_7030 import *
setvars(gene,drug)
from ml_data_7030 import *
# from YC run_all_ML: run locally
#from UQ_yc_RunAllClfs import run_all_ML
#==========================================
# Import ML function: Feature selection
#==========================================
# TT run all ML clfs: feature selection
from FS import fsgs
#==================
# Specify outdir
#==================
outdir_ml = outdir + 'ml/tts_7030/fs/'
print('\nOutput directory:', outdir_ml)
#OutFileFS = outdir_ml + gene.lower() + '_FS' + OutFile_suffix + '.json'
OutFileFS = outdir_ml + gene.lower() + '_FS_noOR' + OutFile_suffix + '.json'
############################################################################
###############################################################################
#====================
# single model CALL
#====================
# aFS = fsgs(input_df = X
# , target = y
# , param_gridLd = [{'fs__min_features_to_select': [1]}]
# , blind_test_df = X_bts
# , blind_test_target = y_bts
# , estimator = LogisticRegression(**rs)
# , use_fs = False # uses estimator as the RFECV parameter for fs. Set to TRUE if you want to supply custom_fs as shown below
# , custom_fs = RFECV(DecisionTreeClassifier(**rs) , cv = skf_cv, scoring = 'matthews_corrcoef')
# , cv_method = skf_cv
# , var_type = 'mixed'
# )
#############
# Loop
############
# models_all = [
# ('XGBoost' , XGBClassifier(**rs, **njobs
# , n_estimators = 100 # wasn't there
# , max_depyth = 3 # wasn't there
# , verbosity = 3
# #, use_label_encoder = False)
# ) )
# ]
models = [('AdaBoost Classifier' , AdaBoostClassifier(**rs) )
##, ('Bagging Classifier' , BaggingClassifier(**rs, **njobs, bootstrap = True, oob_score = True) )
, ('Decision Tree' , DecisionTreeClassifier(**rs) )
, ('Extra Tree' , ExtraTreeClassifier(**rs) )
, ('Extra Trees' , ExtraTreesClassifier(**rs) )
, ('Gradient Boosting' , GradientBoostingClassifier(**rs) )
##, ('Gaussian NB' , GaussianNB() )
##, ('Gaussian Process' , GaussianProcessClassifier(**rs) )
##, ('K-Nearest Neighbors' , KNeighborsClassifier() )
, ('LDA' , LinearDiscriminantAnalysis() )
, ('Logistic Regression' , LogisticRegression(**rs) )
, ('Logistic RegressionCV' , LogisticRegressionCV(cv = 3, **rs))
##, ('MLP' , MLPClassifier(max_iter = 500, **rs) )
##, ('Multinomial' , MultinomialNB() )
##, ('Naive Bayes' , BernoulliNB() )
, ('Passive Aggresive' , PassiveAggressiveClassifier(**rs, **njobs) )
##, ('QDA' , QuadraticDiscriminantAnalysis() )
, ('Random Forest' , RandomForestClassifier(**rs, n_estimators = 1000 ) )
, ('Random Forest2' , RandomForestClassifier(min_samples_leaf = 5
, n_estimators = 1000
, bootstrap = True
, oob_score = True
, **njobs
, **rs
, max_features = 'auto') )
, ('Ridge Classifier' , RidgeClassifier(**rs) )
, ('Ridge ClassifierCV' , RidgeClassifierCV(cv = 3) )
##, ('SVC' , SVC(**rs) )
, ('Stochastic GDescent' , SGDClassifier(**rs, **njobs) )
## , ('XGBoost' , XGBClassifier(**rs, **njobs, verbosity = 3
## , use_label_encoder = False) )
]
print('\n#####################################################################'
, '\nRunning Feature Selection using classfication models (n):', len(models)
, '\nGene:' , gene.lower()
, '\nDrug:' , drug
, '\nSplit:' , tts_split
,'\n####################################################################')
for m in models:
print(m)
print('\n====================================================================\n')
out_fsD = {}
index = 1
for model_name, model_fn in models:
print('\nRunning classifier with FS:', index
, '\nModel_name:' , model_name
, '\nModel func:' , model_fn)
#, '\nList of models:', models)
index = index+1
out_fsD[model_name] = fsgs(input_df = X
, target = y
, param_gridLd = [{'fs__min_features_to_select': [1]}]
, blind_test_df = X_bts
, blind_test_target = y_bts
, estimator = model_fn
, use_fs = False # uses estimator as the RFECV parameter for fs. Set to TRUE if you want to supply custom_fs as shown below
, custom_fs = RFECV(DecisionTreeClassifier(**rs) , cv = skf_cv, scoring = 'matthews_corrcoef')
, cv_method = skf_cv
, var_type = 'mixed'
)
out_fsD
#%% Checking results dict
tot_Ditems = sum(len(v) for v in out_fsD.values())
checkL = []
for k, v in out_fsD.items():
l = [len(out_fsD[k])]
checkL = checkL + l
n_sD = len(checkL) # no. of subDicts
l_sD = list(set(checkL)) # length of each subDict
print('\nTotal no.of subdicts:', n_sD)
if len(l_sD) == 1 and tot_Ditems == n_sD*l_sD[0]:
print('\nPASS: successful run for all Classifiers'
, '\nLength of each subdict:', l_sD)
print('\nSuccessfully ran Feature selection on', len(models), 'classifiers'
, '\nGene:', gene.lower()
, '\nDrug:', drug
, '\nSplit type:', tts_split
, '\nTotal fs models results:', len(out_fsD)
, '\nTotal items in output:', sum(len(v) for v in out_fsD.values()) )
##############################################################################
#%% json output
#========================================
# Write final output file
# https://stackoverflow.com/questions/19201290/how-to-save-a-dictionary-to-a-file
#========================================
# Output final dict as a json
print('\nWriting Final output file (json):', OutFileFS)
with open(OutFileFS, 'w') as f:
f.write(json.dumps(out_fsD
# , cls = NpEncoder
))
# read json
with open(OutFileFS, 'r') as f:data = json.load(f)
##############################################################################

242
scripts/ml/run_FS_7030.py
View file

@ -1,242 +0,0 @@
#!/usr/bin/env python3
# -*- coding: utf-8 -*-
"""
Created on Tue May 24 08:11:05 2022
@author: tanu
"""
#%%
import os, sys
import pandas as pd
import numpy as np
import pprint as pp
from copy import deepcopy
from sklearn import linear_model
from sklearn import datasets
from collections import Counter
from sklearn.linear_model import LogisticRegression, LogisticRegressionCV
from sklearn.linear_model import RidgeClassifier, RidgeClassifierCV, SGDClassifier, PassiveAggressiveClassifier
from sklearn.naive_bayes import BernoulliNB
from sklearn.neighbors import KNeighborsClassifier
from sklearn.svm import SVC
from sklearn.tree import DecisionTreeClassifier, ExtraTreeClassifier
from sklearn.ensemble import RandomForestClassifier, ExtraTreesClassifier, AdaBoostClassifier, GradientBoostingClassifier, BaggingClassifier
from sklearn.naive_bayes import GaussianNB
from sklearn.gaussian_process import GaussianProcessClassifier, kernels
from sklearn.gaussian_process.kernels import RBF, DotProduct, Matern, RationalQuadratic, WhiteKernel
from sklearn.discriminant_analysis import LinearDiscriminantAnalysis, QuadraticDiscriminantAnalysis
from sklearn.neural_network import MLPClassifier
from sklearn.svm import SVC
from xgboost import XGBClassifier
from sklearn.naive_bayes import MultinomialNB
from sklearn.preprocessing import StandardScaler, MinMaxScaler, OneHotEncoder
from sklearn.compose import ColumnTransformer
from sklearn.compose import make_column_transformer
from sklearn.metrics import make_scorer, confusion_matrix, accuracy_score, balanced_accuracy_score, precision_score, average_precision_score, recall_score
from sklearn.metrics import roc_auc_score, roc_curve, f1_score, matthews_corrcoef, jaccard_score, classification_report
# added
from sklearn.model_selection import train_test_split, cross_validate, cross_val_score, LeaveOneOut, KFold, RepeatedKFold, cross_val_predict
from sklearn.model_selection import train_test_split, cross_validate, cross_val_score
from sklearn.model_selection import StratifiedKFold,RepeatedStratifiedKFold, RepeatedKFold
from sklearn.pipeline import Pipeline, make_pipeline
from sklearn.feature_selection import RFE, RFECV
import itertools
import seaborn as sns
import matplotlib.pyplot as plt
from statistics import mean, stdev, median, mode
from imblearn.over_sampling import RandomOverSampler
from imblearn.under_sampling import RandomUnderSampler
from imblearn.over_sampling import SMOTE
from sklearn.datasets import make_classification
from imblearn.combine import SMOTEENN
from imblearn.combine import SMOTETomek
from imblearn.over_sampling import SMOTENC
from imblearn.under_sampling import EditedNearestNeighbours
from imblearn.under_sampling import RepeatedEditedNearestNeighbours
from sklearn.model_selection import GridSearchCV
from sklearn.base import BaseEstimator
from sklearn.impute import KNNImputer as KNN
import json
import argparse
import re
###############################################################################
#gene = 'pncA'
#drug = 'pyrazinamide'
#total_mtblineage_uc = 8
#%% command line args: case sensitive
arg_parser = argparse.ArgumentParser()
arg_parser.add_argument('-d', '--drug', help = 'drug name', default = '')
arg_parser.add_argument('-g', '--gene', help = 'gene name', default = '')
args = arg_parser.parse_args()
drug = args.drug
gene = args.gene
###############################################################################
#==================
# other vars
#==================
tts_split = '70_30'
OutFile_suffix = '7030_FS'
###############################################################################
homedir = os.path.expanduser("~")
sys.path.append(homedir + '/git/LSHTM_analysis/scripts/ml')
###############################################################################
#==================
# Import data
#==================
from ml_data_7030 import *
setvars(gene,drug)
from ml_data_7030 import *
# from YC run_all_ML: run locally
#from UQ_yc_RunAllClfs import run_all_ML
#==========================================
# Import ML functions:
# fsgs_rfecv(): RFECV for Feature selection
#==========================================
from MultClfs import *
#==================
# Specify outdir
#==================
outdir_ml = outdir + 'ml/tts_7030/fs/'
print('\nOutput directory:', outdir_ml)
#OutFileFS = outdir_ml + gene.lower() + '_FS' + OutFile_suffix + '.json'
OutFileFS = outdir_ml + gene.lower() + '_FS_noOR' + OutFile_suffix + '.json'
############################################################################
###############################################################################
#====================
# single model CALL
#====================
# aFS = fsgs(input_df = X
# , target = y
# , param_gridLd = [{'fs__min_features_to_select': [1]}]
# , blind_test_df = X_bts
# , blind_test_target = y_bts
# , estimator = LogisticRegression(**rs)
# , use_fs = False # uses estimator as the RFECV parameter for fs. Set to TRUE if you want to supply custom_fs as shown below
# , custom_fs = RFECV(DecisionTreeClassifier(**rs) , cv = skf_cv, scoring = 'matthews_corrcoef')
# , cv_method = skf_cv
# , var_type = 'mixed'
# )
#############
# Loop
############
#models_fs = [('Decision Tree' , DecisionTreeClassifier(**rs)) ]
models_fs = [('AdaBoost Classifier' , AdaBoostClassifier(**rs) )
, ('Decision Tree' , DecisionTreeClassifier(**rs) )
, ('Extra Tree' , ExtraTreeClassifier(**rs) )
, ('Extra Trees' , ExtraTreesClassifier(**rs) )
, ('Gradient Boosting' , GradientBoostingClassifier(**rs) )
, ('LDA' , LinearDiscriminantAnalysis() )
, ('Logistic Regression' , LogisticRegression(**rs) )
, ('Logistic RegressionCV' , LogisticRegressionCV(cv = 3, **rs))
, ('Passive Aggresive' , PassiveAggressiveClassifier(**rs, **njobs) )
, ('Random Forest' , RandomForestClassifier(**rs, n_estimators = 1000 ) )
, ('Random Forest2' , RandomForestClassifier(min_samples_leaf = 5
, n_estimators = 1000
, bootstrap = True
, oob_score = True
, **njobs
, **rs
, max_features = 'auto') )
, ('Ridge Classifier' , RidgeClassifier(**rs) )
, ('Ridge ClassifierCV' , RidgeClassifierCV(cv = 3) )
, ('Stochastic GDescent' , SGDClassifier(**rs, **njobs) )
## , ('XGBoost' , XGBClassifier(**rs, **njobs, verbosity = 3 , use_label_encoder = False) )
]
print('\n#####################################################################'
, '\nRunning Feature Selection using classfication models_fs (n):', len(models_fs)
, '\nGene:' , gene.lower()
, '\nDrug:' , drug
, '\nSplit:' , tts_split
,'\n####################################################################')
for m in models_fs:
print(m)
print('\n====================================================================\n')
out_fsD = {}
index = 1
for model_name, model_fn in models_fs:
print('\nRunning classifier with FS:', index
, '\nModel_name:' , model_name
, '\nModel func:' , model_fn)
#, '\nList of models_fs:', models_fs)
index = index+1
out_fsD[model_name] = fsgs_rfecv(input_df = X
, target = y
, param_gridLd = [{'fs__min_features_to_select': [1]}]
, blind_test_df = X_bts
, blind_test_target = y_bts
, estimator = model_fn
, use_fs = False # uses estimator as the RFECV parameter for fs. Set to TRUE if you want to supply custom_fs as shown below
, custom_fs = RFECV(DecisionTreeClassifier(**rs) , cv = skf_cv, scoring = 'matthews_corrcoef')
, cv_method = skf_cv
, var_type = 'mixed'
)
out_fsD
#%% Checking results dict
tot_Ditems = sum(len(v) for v in out_fsD.values())
checkL = []
for k, v in out_fsD.items():
l = [len(out_fsD[k])]
checkL = checkL + l
n_sD = len(checkL) # no. of subDicts
l_sD = list(set(checkL)) # length of each subDict
print('\nTotal no.of subdicts:', n_sD)
if len(l_sD) == 1 and tot_Ditems == n_sD*l_sD[0]:
print('\nPASS: successful run for all Classifiers'
, '\nLength of each subdict:', l_sD)
print('\nSuccessfully ran Feature selection on', len(models_fs), 'classifiers'
, '\nGene:', gene.lower()
, '\nDrug:', drug
, '\nSplit type:', tts_split
, '\nTotal fs models results:', len(out_fsD)
, '\nTotal items in output:', sum(len(v) for v in out_fsD.values()) )
##############################################################################
#%% json output
#========================================
# Write final output file
# https://stackoverflow.com/questions/19201290/how-to-save-a-dictionary-to-a-file
#========================================
# Output final dict as a json
print('\nWriting Final output file (json):', OutFileFS)
with open(OutFileFS, 'w') as f:
f.write(json.dumps(out_fsD
# , cls = NpEncoder
))
# read json
with open(OutFileFS, 'r') as f:data = json.load(f)
#############################################################################

142
scripts/ml/run_cd_7030.py
View file

@ -1,142 +0,0 @@
#!/usr/bin/env python3
# -*- coding: utf-8 -*-
"""
Created on Mon Jun 20 13:05:23 2022
@author: tanu
"""
#%%Imports ####################################################################
import re
import argparse
import os, sys
# gene = 'pncA'
# drug = 'pyrazinamide'
#total_mtblineage_uc = 8
###############################################################################
#%% command line args: case sensitive
arg_parser = argparse.ArgumentParser()
arg_parser.add_argument('-d', '--drug', help = 'drug name', default = '')
arg_parser.add_argument('-g', '--gene', help = 'gene name', default = '')
args = arg_parser.parse_args()
drug = args.drug
gene = args.gene
###############################################################################
homedir = os.path.expanduser("~")
sys.path.append(homedir + '/git/LSHTM_analysis/scripts/ml')
###############################################################################
#==================
# Import data
#==================
from ml_data_cd_7030 import *
setvars(gene,drug)
from ml_data_cd_7030 import *
# from YC run_all_ML: run locally
#from UQ_yc_RunAllClfs import run_all_ML
#====================
# Import ML functions
#====================
from MultClfs import *
#==================
# other vars
#==================
tts_split_cd_7030 = 'cd_7030'
OutFile_suffix = '_cd_7030'
#==================
# Specify outdir
#==================
outdir_ml = outdir + 'ml/tts_cd_7030/'
print('\nOutput directory:', outdir_ml)
#outFile_wf = outdir_ml + gene.lower() + '_baselineC_' + OutFile_suffix + '.csv'
outFile_wf = outdir_ml + gene.lower() + '_baselineC_noOR' + OutFile_suffix + '.csv'
#%% Running models ############################################################
print('\n#####################################################################\n'
, '\nStarting--> Running ML analysis: Baseline modes (No FS)'
, '\nGene name:', gene
, '\nDrug name:', drug
, '\n#####################################################################\n')
paramD = {
'baseline_paramD': { 'input_df' : X
, 'target' : y
, 'var_type' : 'mixed'
, 'resampling_type': 'none'}
, 'smnc_paramD': { 'input_df' : X_smnc
, 'target' : y_smnc
, 'var_type' : 'mixed'
, 'resampling_type' : 'smnc'}
, 'ros_paramD': { 'input_df' : X_ros
, 'target' : y_ros
, 'var_type' : 'mixed'
, 'resampling_type' : 'ros'}
, 'rus_paramD' : { 'input_df' : X_rus
, 'target' : y_rus
, 'var_type' : 'mixed'
, 'resampling_type' : 'rus'}
, 'rouC_paramD' : { 'input_df' : X_rouC
, 'target' : y_rouC
, 'var_type' : 'mixed'
, 'resampling_type' : 'rouC'}
}
##==============================================================================
## Dict with no CV BT formatted df
## mmD = {}
## for k, v in paramD.items():
## # print(mmD[k])
## scores_cd_7030D = MultModelsCl(**paramD[k]
## , tts_split_type = tts_split_cd_7030
## , skf_cv = skf_cv
## , blind_test_df = X_bts
## , blind_test_target = y_bts
## , add_cm = True
## , add_yn = True
## , return_formatted_output = False)
## mmD[k] = scores_cd_7030D
##==============================================================================
## Initial run to get the dict of dicts for each sampling type containing CV, BT and metadata DFs
mmDD = {}
for k, v in paramD.items():
scores_cd_7030D = MultModelsCl(**paramD[k]
, tts_split_type = tts_split_cd_7030
, skf_cv = skf_cv
, blind_test_df = X_bts
, blind_test_target = y_bts
, add_cm = True
, add_yn = True
, return_formatted_output = True)
mmDD[k] = scores_cd_7030D
# Extracting the dfs from within the dict and concatenating to output as one df
for k, v in mmDD.items():
out_wf_cd_7030 = pd.concat(mmDD, ignore_index = True)
out_wf_cd_7030f = out_wf_cd_7030.sort_values(by = ['resampling', 'source_data', 'MCC'], ascending = [True, True, False], inplace = False)
print('\n######################################################################'
, '\nEnd--> Successfully generated output DF for Multiple classifiers (baseline models)'
, '\nGene:', gene.lower()
, '\nDrug:', drug
, '\noutput file:', outFile_wf
, '\nDim of output:', out_wf_cd_7030f.shape
, '\n######################################################################')
###############################################################################
#====================
# Write output file
#====================
out_wf_cd_7030f.to_csv(outFile_wf, index = False)
print('\nFile successfully written:', outFile_wf)
###############################################################################

141
scripts/ml/run_cd_8020.py
View file

@ -1,141 +0,0 @@
#!/usr/bin/env python3
# -*- coding: utf-8 -*-
"""
Created on Mon Jun 20 13:05:23 2022
@author: tanu
"""
#%%Imports ####################################################################
import re
import argparse
import os, sys
# gene = 'pncA'
# drug = 'pyrazinamide'
#total_mtblineage_uc = 8
###############################################################################
#%% command line args: case sensitive
arg_parser = argparse.ArgumentParser()
arg_parser.add_argument('-d', '--drug', help = 'drug name', default = '')
arg_parser.add_argument('-g', '--gene', help = 'gene name', default = '')
args = arg_parser.parse_args()
drug = args.drug
gene = args.gene
###############################################################################
homedir = os.path.expanduser("~")
sys.path.append(homedir + '/git/LSHTM_analysis/scripts/ml')
###############################################################################
#==================
# Import data
#==================
from ml_data_cd_8020 import *
setvars(gene,drug)
from ml_data_cd_8020 import *
# from YC run_all_ML: run locally
#from UQ_yc_RunAllClfs import run_all_ML
#====================
# Import ML functions
#====================
from MultClfs import *
#==================
# other vars
#==================
tts_split_cd_8020 = 'cd_80_20'
OutFile_suffix = '_cd_8020'
#==================
# Specify outdir
#==================
outdir_ml = outdir + 'ml/tts_cd_8020/'
print('\nOutput directory:', outdir_ml)
#outFile_wf = outdir_ml + gene.lower() + '_baselineC_' + OutFile_suffix + '.csv'
outFile_wf = outdir_ml + gene.lower() + '_baselineC_noOR' + OutFile_suffix + '.csv'
#%% Running models ############################################################
print('\n#####################################################################\n'
, '\nStarting--> Running ML analysis: Baseline modes (No FS)'
, '\nGene name:', gene
, '\nDrug name:', drug
, '\n#####################################################################\n')
paramD = {
'baseline_paramD': { 'input_df' : X
, 'target' : y
, 'var_type' : 'mixed'
, 'resampling_type': 'none'}
, 'smnc_paramD': { 'input_df' : X_smnc
, 'target' : y_smnc
, 'var_type' : 'mixed'
, 'resampling_type' : 'smnc'}
, 'ros_paramD': { 'input_df' : X_ros
, 'target' : y_ros
, 'var_type' : 'mixed'
, 'resampling_type' : 'ros'}
, 'rus_paramD' : { 'input_df' : X_rus
, 'target' : y_rus
, 'var_type' : 'mixed'
, 'resampling_type' : 'rus'}
, 'rouC_paramD' : { 'input_df' : X_rouC
, 'target' : y_rouC
, 'var_type' : 'mixed'
, 'resampling_type' : 'rouC'}
}
##==============================================================================
## Dict with no CV BT formatted df
## mmD = {}
## for k, v in paramD.items():
## # print(mmD[k])
## scores_cd_8020D = MultModelsCl(**paramD[k]
## , tts_split_type = tts_split_cd_8020
## , skf_cv = skf_cv
## , blind_test_df = X_bts
## , blind_test_target = y_bts
## , add_cm = True
## , add_yn = True
## , return_formatted_output = False)
## mmD[k] = scores_cd_8020D
##==============================================================================
## Initial run to get the dict of dicts for each sampling type containing CV, BT and metadata DFs
mmDD = {}
for k, v in paramD.items():
scores_cd_8020D = MultModelsCl(**paramD[k]
, tts_split_type = tts_split_cd_8020
, skf_cv = skf_cv
, blind_test_df = X_bts
, blind_test_target = y_bts
, add_cm = True
, add_yn = True
, return_formatted_output = True)
mmDD[k] = scores_cd_8020D
# Extracting the dfs from within the dict and concatenating to output as one df
for k, v in mmDD.items():
out_wf_cd_8020 = pd.concat(mmDD, ignore_index = True)
out_wf_cd_8020f = out_wf_cd_8020.sort_values(by = ['resampling', 'source_data', 'MCC'], ascending = [True, True, False], inplace = False)
print('\n######################################################################'
, '\nEnd--> Successfully generated output DF for Multiple classifiers (baseline models)'
, '\nGene:', gene.lower()
, '\nDrug:', drug
, '\noutput file:', outFile_wf
, '\nDim of output:', out_wf_cd_8020f.shape
, '\n######################################################################')
###############################################################################
#====================
# Write output file
#====================
out_wf_cd_8020f.to_csv(outFile_wf, index = False)
print('\nFile successfully written:', outFile_wf)
###############################################################################

141
scripts/ml/run_cd_sl.py
View file

@ -1,141 +0,0 @@
#!/usr/bin/env python3
# -*- coding: utf-8 -*-
"""
Created on Mon Jun 20 13:05:23 2022
@author: tanu
"""
#%%Imports ####################################################################
import re
import argparse
import os, sys
# gene = 'pncA'
# drug = 'pyrazinamide'
#total_mtblineage_uc = 8
###############################################################################
#%% command line args: case sensitive
arg_parser = argparse.ArgumentParser()
arg_parser.add_argument('-d', '--drug', help = 'drug name', default = '')
arg_parser.add_argument('-g', '--gene', help = 'gene name', default = '')
args = arg_parser.parse_args()
drug = args.drug
gene = args.gene
###############################################################################
homedir = os.path.expanduser("~")
sys.path.append(homedir + '/git/LSHTM_analysis/scripts/ml')
###############################################################################
#==================
# Import data
#==================
from ml_data_cd_sl import *
setvars(gene,drug)
from ml_data_cd_sl import *
# from YC run_all_ML: run locally
#from UQ_yc_RunAllClfs import run_all_ML
#====================
# Import ML functions
#====================
from MultClfs import *
#==================
# other vars
#==================
tts_split_cd_sl = 'cd_sl'
OutFile_suffix = '_cd_sl'
#==================
# Specify outdir
#==================
outdir_ml = outdir + 'ml/tts_cd_sl/'
print('\nOutput directory:', outdir_ml)
#outFile_wf = outdir_ml + gene.lower() + '_baselineC_' + OutFile_suffix + '.csv'
outFile_wf = outdir_ml + gene.lower() + '_baselineC_noOR' + OutFile_suffix + '.csv'
#%% Running models ############################################################
print('\n#####################################################################\n'
, '\nStarting--> Running ML analysis: Baseline modes (No FS)'
, '\nGene name:', gene
, '\nDrug name:', drug
, '\n#####################################################################\n')
paramD = {
'baseline_paramD': { 'input_df' : X
, 'target' : y
, 'var_type' : 'mixed'
, 'resampling_type': 'none'}
, 'smnc_paramD': { 'input_df' : X_smnc
, 'target' : y_smnc
, 'var_type' : 'mixed'
, 'resampling_type' : 'smnc'}
, 'ros_paramD': { 'input_df' : X_ros
, 'target' : y_ros
, 'var_type' : 'mixed'
, 'resampling_type' : 'ros'}
, 'rus_paramD' : { 'input_df' : X_rus
, 'target' : y_rus
, 'var_type' : 'mixed'
, 'resampling_type' : 'rus'}
, 'rouC_paramD' : { 'input_df' : X_rouC
, 'target' : y_rouC
, 'var_type' : 'mixed'
, 'resampling_type' : 'rouC'}
}
##==============================================================================
## Dict with no CV BT formatted df
## mmD = {}
## for k, v in paramD.items():
## # print(mmD[k])
## scores_cd_slD = MultModelsCl(**paramD[k]
## , tts_split_type = tts_split_cd_sl
## , skf_cv = skf_cv
## , blind_test_df = X_bts
## , blind_test_target = y_bts
## , add_cm = True
## , add_yn = True
## , return_formatted_output = False)
## mmD[k] = scores_cd_slD
##==============================================================================
## Initial run to get the dict of dicts for each sampling type containing CV, BT and metadata DFs
mmDD = {}
for k, v in paramD.items():
scores_cd_slD = MultModelsCl(**paramD[k]
, tts_split_type = tts_split_cd_sl
, skf_cv = skf_cv
, blind_test_df = X_bts
, blind_test_target = y_bts
, add_cm = True
, add_yn = True
, return_formatted_output = True)
mmDD[k] = scores_cd_slD
# Extracting the dfs from within the dict and concatenating to output as one df
for k, v in mmDD.items():
out_wf_cd_sl = pd.concat(mmDD, ignore_index = True)
out_wf_cd_slf = out_wf_cd_sl.sort_values(by = ['resampling', 'source_data', 'MCC'], ascending = [True, True, False], inplace = False)
print('\n######################################################################'
, '\nEnd--> Successfully generated output DF for Multiple classifiers (baseline models)'
, '\nGene:', gene.lower()
, '\nDrug:', drug
, '\noutput file:', outFile_wf
, '\nDim of output:', out_wf_cd_slf.shape
, '\n######################################################################')
###############################################################################
#====================
# Write output file
#====================
out_wf_cd_slf.to_csv(outFile_wf, index = False)
print('\nFile successfully written:', outFile_wf)
###############################################################################

557
scripts/ml/run_fg.py
View file

@ -1,557 +0,0 @@
#!/usr/bin/env python3
# -*- coding: utf-8 -*-
"""
Created on Sat May 28 05:25:30 2022
@author: tanu
"""
import os
import re
import argparse
###############################################################################
# gene = 'pncA'
# drug = 'pyrazinamide'
#total_mtblineage_uc = 8
#%% command line args: case sensitive
arg_parser = argparse.ArgumentParser()
arg_parser.add_argument('-d', '--drug', help = 'drug name', default = '')
arg_parser.add_argument('-g', '--gene', help = 'gene name', default = '')
args = arg_parser.parse_args()
drug = args.drug
gene = args.gene
###############################################################################
homedir = os.path.expanduser("~")
os.chdir( homedir + '/git/LSHTM_analysis/scripts/ml/')
#==================
# Import data
#==================
from ml_data_fg import *
setvars(gene,drug)
from ml_data_fg import *
# from YC run_all_ML: run locally
#from UQ_yc_RunAllClfs import run_all_ML
#====================
# Import ML function
#====================
# TT run all ML clfs: baseline model
from MultModelsCl import MultModelsCl
############################################################################
print('\n#####################################################################\n'
, '\nRunning ML analysis: feature groups '
, '\nGene name:', gene
, '\nDrug name:', drug)
#==================
# Specify outdir
#==================
outdir_ml = outdir + 'ml/uq_v1/fgs/'
print('\nOutput directory:', outdir_ml)
outFile = outdir_ml + gene.lower() + '_baseline_FG.csv'
#==================
# other vars
#==================
tts_split = 'original'
resampling = 'none'
###############################################################################
score_type_ordermapD = { 'mcc' : 1
, 'fscore' : 2
, 'jcc' : 3
, 'precision' : 4
, 'recall' : 5
, 'accuracy' : 6
, 'roc_auc' : 7
, 'TN' : 8
, 'FP' : 9
, 'FN' : 10
, 'TP' : 11
, 'trainingY_neg': 12
, 'trainingY_pos': 13
, 'blindY_neg' : 14
, 'blindY_pos' : 15
, 'fit_time' : 16
, 'score_time' : 17
}
#%%###########################################################################
print('\n================================================================\n')
#all_featuresN = X_evolFN + X_structural_FN + X_genomicFN
# X_structural_FN = X_stability_FN + X_affinityFN + X_resprop_FN
# X_resprop_FN = X_aaindex_Fnum + X_str_Fnum + X_aap_Fcat
print('\n================================================================'
, '\nTotal Evolutionary features (n):' , len(X_evolFN)
, '\n--------------Evol. feature colnames:', X_evolFN
, '\n================================================================'
, '\n\nTotal structural features (n):', len(X_structural_FN)
, '\n--------Stability ncols:' , len(X_stability_FN)
, '\n--------------Common stability colnames:' , X_common_stability_Fnum
, '\n--------------Foldx colnames:' , X_foldX_Fnum
, '\n--------Affinity ncols:' , len(X_affinityFN)
, '\n--------------Common affinity colnames:' , common_affinity_Fnum
, '\n--------------Gene specific affinity colnames:', gene_affinity_colnames
, '\n--------Residue prop ncols:' , len(X_resprop_FN)
, '\n--------------Residue Prop cols:' , X_str_Fnum
, '\n--------------AA change Prop cols:' , X_aap_Fcat
, '\n--------------AA index cols:' , X_aaindex_Fnum
, '\n================================================================'
, '\n\nTotal Genomic features (n):' , len(X_genomicFN)
, '\n--------MAF+OR cols:' , len(X_gn_mafor_Fnum)
, '\n--------------MAF+OR colnames:' , X_gn_mafor_Fnum
, '\n--------Lineage cols:' , len(X_gn_linegae_Fnum)
, '\n--------------Lineage cols:' , X_gn_linegae_Fnum
, '\n--------Other cols:' , len(X_gn_Fcat)
, '\n--------------Other cols:' , X_gn_Fcat
, '\n================================================================')
# Sanity check
if ( len(X.columns) == len(X_evolFN) + len(X_structural_FN) + len(X_genomicFN)):
print('\nPass: No. of features match')
else:
print('\nFail: Count of feature mismatch'
, '\nExpected:', len(X_evolFN) + len(X_structural_FN) + len(X_genomicFN)
, '\nGot:', len(X.columns))
sys.exit()
print('\n#####################################################################\n')
###############################################################################
#================
# Evolutionary
# X_evolFN
#================
feature_gp_nameEV = 'evolutionary'
n_featuresEV = len(X_evolFN)
scores_mmEV = MultModelsCl(input_df = X[X_evolFN]
, target = y
, var_type = 'mixed'
, skf_cv = skf_cv
, blind_test_input_df = X_bts[X_evolFN]
, blind_test_target = y_bts
, add_cm = True
, add_yn = True)
baseline_allEV = pd.DataFrame(scores_mmEV)
baseline_EV = baseline_allEV.filter(regex = 'bts_.*|test_.*|.*_time|TN|FP|FN|TP|.*_neg|.*_pos', axis = 0)
baseline_EV = baseline_EV.reset_index()
baseline_EV.rename(columns = {'index': 'original_names'}, inplace = True)
# Indicate whether BT or CT
bt_pattern = re.compile(r'bts_.*')
baseline_EV['data_source'] = baseline_EV.apply(lambda row: 'BT' if bt_pattern.search(row.original_names) else 'CV' , axis = 1)
baseline_EV['score_type'] = baseline_EV['original_names'].str.replace('bts_|test_', '', regex = True)
score_type_uniqueN = set(baseline_EV['score_type'])
cL1 = list(score_type_ordermapD.keys())
cL2 = list(score_type_uniqueN)
if set(cL1).issubset(cL2):
print('\nPASS: sorting df by score that is mapped onto the order I want')
baseline_EV['score_order'] = baseline_EV['score_type'].map(score_type_ordermapD)
baseline_EV.sort_values(by = ['data_source', 'score_order'], ascending = [True, True], inplace = True)
else:
sys.exit('\nFAIL: could not sort df as score mapping for ordering failed')
baseline_EV['feature_group'] = feature_gp_nameEV
baseline_EV['resampling'] = resampling
baseline_EV['tts_split'] = tts_split
baseline_EV['n_features'] = n_featuresEV
###############################################################################
#================
# Genomics
# X_genomicFN
#================
feature_gp_nameGN = 'genomics'
n_featuresGN = len(X_genomicFN)
scores_mmGN = MultModelsCl(input_df = X[X_genomicFN]
, target = y
, var_type = 'mixed'
, skf_cv = skf_cv
, blind_test_input_df = X_bts[X_genomicFN]
, blind_test_target = y_bts
, add_cm = True
, add_yn = True)
baseline_allGN = pd.DataFrame(scores_mmGN)
baseline_GN = baseline_allGN.filter(regex = 'bts_.*|test_.*|.*_time|TN|FP|FN|TP|.*_neg|.*_pos', axis = 0)
baseline_GN = baseline_GN.reset_index()
baseline_GN.rename(columns = {'index': 'original_names'}, inplace = True)
# Indicate whether BT or CT
bt_pattern = re.compile(r'bts_.*')
baseline_GN['data_source'] = baseline_GN.apply(lambda row: 'BT' if bt_pattern.search(row.original_names) else 'CV' , axis = 1)
baseline_GN['score_type'] = baseline_GN['original_names'].str.replace('bts_|test_', '', regex = True)
score_type_uniqueN = set(baseline_GN['score_type'])
cL1 = list(score_type_ordermapD.keys())
cL2 = list(score_type_uniqueN)
if set(cL1).issubset(cL2):
print('\nPASS: sorting df by score that is mapped onto the order I want')
baseline_GN['score_order'] = baseline_GN['score_type'].map(score_type_ordermapD)
baseline_GN.sort_values(by = ['data_source', 'score_order'], ascending = [True, True], inplace = True)
else:
sys.exit('\nFAIL: could not sort df as score mapping for ordering failed')
baseline_GN['feature_group'] = feature_gp_nameGN
baseline_GN['resampling'] = resampling
baseline_GN['tts_split'] = tts_split
baseline_GN['n_features'] = n_featuresGN
###############################################################################
#all_featuresN = X_evolFN + X_structural_FN + X_genomicFN
# X_structural_FN = X_stability_FN + X_affinityFN + X_resprop_FN
# X_resprop_FN = X_aaindex_Fnum + X_str_Fnum + X_aap_Fcat
#================
# Structural cols
# X_structural_FN
#================
feature_gp_nameSTR = 'structural'
n_featuresSTR = len(X_structural_FN)
scores_mmSTR = MultModelsCl(input_df = X[X_structural_FN]
, target = y
, var_type = 'mixed'
, skf_cv = skf_cv
, blind_test_input_df = X_bts[X_structural_FN]
, blind_test_target = y_bts
, add_cm = True
, add_yn = True)
baseline_allSTR = pd.DataFrame(scores_mmSTR)
baseline_STR = baseline_allSTR.filter(regex = 'bts_.*|test_.*|.*_time|TN|FP|FN|TP|.*_neg|.*_pos', axis = 0)
baseline_STR = baseline_STR.reset_index()
baseline_STR.rename(columns = {'index': 'original_names'}, inplace = True)
# Indicate whether BT or CT
bt_pattern = re.compile(r'bts_.*')
baseline_STR['data_source'] = baseline_STR.apply(lambda row: 'BT' if bt_pattern.search(row.original_names) else 'CV' , axis = 1)
baseline_STR['score_type'] = baseline_STR['original_names'].str.replace('bts_|test_', '', regex = True)
score_type_uniqueN = set(baseline_STR['score_type'])
cL1 = list(score_type_ordermapD.keys())
cL2 = list(score_type_uniqueN)
if set(cL1).issubset(cL2):
print('\nPASS: sorting df by score that is mapped onto the order I want')
baseline_STR['score_order'] = baseline_STR['score_type'].map(score_type_ordermapD)
baseline_STR.sort_values(by = ['data_source', 'score_order'], ascending = [True, True], inplace = True)
else:
sys.exit('\nFAIL: could not sort df as score mapping for ordering failed')
baseline_STR['feature_group'] = feature_gp_nameSTR
baseline_STR['resampling'] = resampling
baseline_STR['tts_split'] = tts_split
baseline_STR['n_features'] = n_featuresSTR
##############################################################################
#================
# Stability cols
# X_stability_FN
#================
feature_gp_nameSTB = 'stability'
n_featuresSTB = len(X_stability_FN)
scores_mmSTB = MultModelsCl(input_df = X[X_stability_FN]
, target = y
, var_type = 'mixed'
, skf_cv = skf_cv
, blind_test_input_df = X_bts[X_stability_FN]
, blind_test_target = y_bts
, add_cm = True
, add_yn = True)
baseline_allSTB = pd.DataFrame(scores_mmSTB)
baseline_STB = baseline_allSTB.filter(regex = 'bts_.*|test_.*|.*_time|TN|FP|FN|TP|.*_neg|.*_pos', axis = 0)
baseline_STB = baseline_STB.reset_index()
baseline_STB.rename(columns = {'index': 'original_names'}, inplace = True)
# Indicate whether BT or CT
bt_pattern = re.compile(r'bts_.*')
baseline_STB['data_source'] = baseline_STB.apply(lambda row: 'BT' if bt_pattern.search(row.original_names) else 'CV' , axis = 1)
baseline_STB['score_type'] = baseline_STB['original_names'].str.replace('bts_|test_', '', regex = True)
score_type_uniqueN = set(baseline_STB['score_type'])
cL1 = list(score_type_ordermapD.keys())
cL2 = list(score_type_uniqueN)
if set(cL1).issubset(cL2):
print('\nPASS: sorting df by score that is mapped onto the order I want')
baseline_STB['score_order'] = baseline_STB['score_type'].map(score_type_ordermapD)
baseline_STB.sort_values(by = ['data_source', 'score_order'], ascending = [True, True], inplace = True)
else:
sys.exit('\nFAIL: could not sort df as score mapping for ordering failed')
baseline_STB['feature_group'] = feature_gp_nameSTB
baseline_STB['resampling'] = resampling
baseline_STB['tts_split'] = tts_split
baseline_STB['n_features'] = n_featuresSTB
###############################################################################
#================
# Affinity cols
# X_affinityFN
#================
feature_gp_nameAFF = 'affinity'
n_featuresAFF = len(X_affinityFN)
scores_mmAFF = MultModelsCl(input_df = X[X_affinityFN]
, target = y
, var_type = 'mixed'
, skf_cv = skf_cv
, blind_test_input_df = X_bts[X_affinityFN]
, blind_test_target = y_bts
, add_cm = True
, add_yn = True)
baseline_allAFF = pd.DataFrame(scores_mmAFF)
baseline_AFF = baseline_allAFF.filter(regex = 'bts_.*|test_.*|.*_time|TN|FP|FN|TP|.*_neg|.*_pos', axis = 0)
baseline_AFF = baseline_AFF.reset_index()
baseline_AFF.rename(columns = {'index': 'original_names'}, inplace = True)
# Indicate whether BT or CT
bt_pattern = re.compile(r'bts_.*')
baseline_AFF['data_source'] = baseline_AFF.apply(lambda row: 'BT' if bt_pattern.search(row.original_names) else 'CV' , axis = 1)
baseline_AFF['score_type'] = baseline_AFF['original_names'].str.replace('bts_|test_', '', regex = True)
score_type_uniqueN = set(baseline_AFF['score_type'])
cL1 = list(score_type_ordermapD.keys())
cL2 = list(score_type_uniqueN)
if set(cL1).issubset(cL2):
print('\nPASS: sorting df by score that is mapped onto the order I want')
baseline_AFF['score_order'] = baseline_AFF['score_type'].map(score_type_ordermapD)
baseline_AFF.sort_values(by = ['data_source', 'score_order'], ascending = [True, True], inplace = True)
else:
sys.exit('\nFAIL: could not sort df as score mapping for ordering failed')
baseline_AFF['feature_group'] = feature_gp_nameAFF
baseline_AFF['resampling'] = resampling
baseline_AFF['tts_split'] = tts_split
baseline_AFF['n_features'] = n_featuresAFF
###############################################################################
#================
# Residue level
# X_resprop_FN
#================
feature_gp_nameRES = 'residue_prop'
n_featuresRES = len(X_resprop_FN)
scores_mmRES = MultModelsCl(input_df = X[X_resprop_FN]
, target = y
, var_type = 'mixed'
, skf_cv = skf_cv
, blind_test_input_df = X_bts[X_resprop_FN]
, blind_test_target = y_bts
, add_cm = True
, add_yn = True)
baseline_allRES = pd.DataFrame(scores_mmRES)
baseline_RES = baseline_allRES.filter(regex = 'bts_.*|test_.*|.*_time|TN|FP|FN|TP|.*_neg|.*_pos', axis = 0)
baseline_RES = baseline_RES.reset_index()
baseline_RES.rename(columns = {'index': 'original_names'}, inplace = True)
# Indicate whether BT or CT
bt_pattern = re.compile(r'bts_.*')
baseline_RES['data_source'] = baseline_RES.apply(lambda row: 'BT' if bt_pattern.search(row.original_names) else 'CV' , axis = 1)
baseline_RES['score_type'] = baseline_RES['original_names'].str.replace('bts_|test_', '', regex = True)
score_type_uniqueN = set(baseline_RES['score_type'])
cL1 = list(score_type_ordermapD.keys())
cL2 = list(score_type_uniqueN)
if set(cL1).issubset(cL2):
print('\nPASS: sorting df by score that is mapped onto the order I want')
baseline_RES['score_order'] = baseline_RES['score_type'].map(score_type_ordermapD)
baseline_RES.sort_values(by = ['data_source', 'score_order'], ascending = [True, True], inplace = True)
else:
sys.exit('\nFAIL: could not sort df as score mapping for ordering failed')
baseline_RES['feature_group'] = feature_gp_nameRES
baseline_RES['resampling'] = resampling
baseline_RES['tts_split'] = tts_split
baseline_RES['n_features'] = n_featuresRES
###############################################################################
#================
# Residue level-AAindex
#X_resprop_FN - X_aaindex_Fnum
#================
X_respropNOaaFN = list(set(X_resprop_FN) - set(X_aaindex_Fnum))
feature_gp_nameRNAA = 'ResPropNoAA'
n_featuresRNAA = len(X_respropNOaaFN)
scores_mmRNAA = MultModelsCl(input_df = X[X_respropNOaaFN]
, target = y
, var_type = 'mixed'
, skf_cv = skf_cv
, blind_test_input_df = X_bts[X_respropNOaaFN]
, blind_test_target = y_bts
, add_cm = True
, add_yn = True)
baseline_allRNAA = pd.DataFrame(scores_mmRNAA)
baseline_RNAA = baseline_allRNAA.filter(regex = 'bts_.*|test_.*|.*_time|TN|FP|FN|TP|.*_neg|.*_pos', axis = 0)
baseline_RNAA = baseline_RNAA.reset_index()
baseline_RNAA.rename(columns = {'index': 'original_names'}, inplace = True)
# Indicate whether BT or CT
bt_pattern = re.compile(r'bts_.*')
baseline_RNAA['data_source'] = baseline_RNAA.apply(lambda row: 'BT' if bt_pattern.search(row.original_names) else 'CV' , axis = 1)
baseline_RNAA['score_type'] = baseline_RNAA['original_names'].str.replace('bts_|test_', '', regex = True)
score_type_uniqueN = set(baseline_RNAA['score_type'])
cL1 = list(score_type_ordermapD.keys())
cL2 = list(score_type_uniqueN)
if set(cL1).issubset(cL2):
print('\nPASS: sorting df by score that is mapped onto the order I want')
baseline_RNAA['score_order'] = baseline_RNAA['score_type'].map(score_type_ordermapD)
baseline_RNAA.sort_values(by = ['data_source', 'score_order'], ascending = [True, True], inplace = True)
else:
sys.exit('\nFAIL: could not sort df as score mapping for ordering failed')
baseline_RNAA['feature_group'] = feature_gp_nameRNAA
baseline_RNAA['resampling'] = resampling
baseline_RNAA['tts_split'] = tts_split
baseline_RNAA['n_features'] = n_featuresRNAA
###############################################################################
#================
# Structural cols-AAindex
#X_structural_FN - X_aaindex_Fnum
#================
X_strNOaaFN = list(set(X_structural_FN) - set(X_aaindex_Fnum))
feature_gp_nameSNAA = 'StrNoAA'
n_featuresSNAA = len(X_strNOaaFN)
scores_mmSNAA = MultModelsCl(input_df = X[X_strNOaaFN]
, target = y
, var_type = 'mixed'
, skf_cv = skf_cv
, blind_test_input_df = X_bts[X_strNOaaFN]
, blind_test_target = y_bts
, add_cm = True
, add_yn = True)
baseline_allSNAA = pd.DataFrame(scores_mmSNAA)
baseline_SNAA = baseline_allSNAA.filter(regex = 'bts_.*|test_.*|.*_time|TN|FP|FN|TP|.*_neg|.*_pos', axis = 0)
baseline_SNAA = baseline_SNAA.reset_index()
baseline_SNAA.rename(columns = {'index': 'original_names'}, inplace = True)
# Indicate whether BT or CT
bt_pattern = re.compile(r'bts_.*')
baseline_SNAA['data_source'] = baseline_SNAA.apply(lambda row: 'BT' if bt_pattern.search(row.original_names) else 'CV' , axis = 1)
baseline_SNAA['score_type'] = baseline_SNAA['original_names'].str.replace('bts_|test_', '', regex = True)
score_type_uniqueN = set(baseline_SNAA['score_type'])
cL1 = list(score_type_ordermapD.keys())
cL2 = list(score_type_uniqueN)
if set(cL1).issubset(cL2):
print('\nPASS: sorting df by score that is mapped onto the order I want')
baseline_SNAA['score_order'] = baseline_SNAA['score_type'].map(score_type_ordermapD)
baseline_SNAA.sort_values(by = ['data_source', 'score_order'], ascending = [True, True], inplace = True)
else:
sys.exit('\nFAIL: could not sort df as score mapping for ordering failed')
baseline_SNAA['feature_group'] = feature_gp_nameSNAA
baseline_SNAA['resampling'] = resampling
baseline_SNAA['tts_split'] = tts_split
baseline_SNAA['n_features'] = n_featuresSNAA
###############################################################################
#%% COMBINING all FG dfs
#================
# Combine all
# https://stackoverflow.com/questions/39862654/pandas-concat-of-multiple-data-frames-using-only-common-columns
#================
dfs_combine = [baseline_EV, baseline_GN, baseline_STR, baseline_STB, baseline_AFF, baseline_RES , baseline_RNAA , baseline_SNAA]
dfs_nrows = []
for df in dfs_combine:
dfs_nrows = dfs_nrows + [len(df)]
dfs_nrows = max(dfs_nrows)
dfs_ncols = []
for df in dfs_combine:
dfs_ncols = dfs_ncols + [len(df.columns)]
dfs_ncols = max(dfs_ncols)
# dfs_ncols = []
# dfs_ncols2 = mode(dfs_ncols.append(len(df.columns) for df in dfs_combine)
# dfs_ncols2
expected_nrows = len(dfs_combine) * dfs_nrows
expected_ncols = dfs_ncols
common_cols = list(set.intersection(*(set(df.columns) for df in dfs_combine)))
if len(common_cols) == dfs_ncols :
combined_FG_baseline = pd.concat([df[common_cols] for df in dfs_combine], ignore_index=True)
fgs = combined_FG_baseline[['feature_group', 'n_features']]
fgs = fgs.drop_duplicates()
print('\nConcatenating dfs with feature groups after ML analysis:'
, '\nNo. of dfs combining:', len(dfs_combine)
, '\nSampling type:', resampling
, '\nThe feature groups are:'
, '\n', fgs)
if len(combined_FG_baseline) == expected_nrows and len(combined_FG_baseline.columns) == expected_ncols:
print('\nPASS:', len(dfs_combine), 'dfs successfully combined'
, '\nnrows in combined_df:', len(combined_FG_baseline)
, '\nncols in combined_df:', len(combined_FG_baseline.columns))
else:
print('\nFAIL: concatenating failed'
, '\nExpected nrows:', expected_nrows
, '\nGot:', len(combined_FG_baseline)
, '\nExpected ncols:', expected_ncols
, '\nGot:', len(combined_FG_baseline.columns))
sys.exit()
else:
sys.exit('\nConcatenting dfs not possible,check numbers ')
# # rpow bind
# if all(ll((baseline_EV.columns == baseline_GN.columns == baseline_STR.columns)):
# print('\nPASS:colnames match, proceeding to rowbind')
# comb_df = pd.concat()], axis = 0, ignore_index = True )
###############################################################################
#====================
# Write output file
#====================
combined_FG_baseline.to_csv(outFile, index = False)
print('\nFile successfully written:', outFile)
###############################################################################

141
scripts/ml/run_sl.py
View file

@ -1,141 +0,0 @@
#!/usr/bin/env python3
# -*- coding: utf-8 -*-
"""
Created on Mon Jun 20 13:05:23 2022
@author: tanu
"""
#%%Imports ####################################################################
import re
import argparse
import os, sys
# gene = 'pncA'
# drug = 'pyrazinamide'
#total_mtblineage_uc = 8
###############################################################################
#%% command line args: case sensitive
arg_parser = argparse.ArgumentParser()
arg_parser.add_argument('-d', '--drug', help = 'drug name', default = '')
arg_parser.add_argument('-g', '--gene', help = 'gene name', default = '')
args = arg_parser.parse_args()
drug = args.drug
gene = args.gene
###############################################################################
homedir = os.path.expanduser("~")
sys.path.append(homedir + '/git/LSHTM_analysis/scripts/ml')
###############################################################################
#==================
# Import data
#==================
from ml_data_sl import *
setvars(gene,drug)
from ml_data_sl import *
# from YC run_all_ML: run locally
#from UQ_yc_RunAllClfs import run_all_ML
#====================
# Import ML functions
#====================
from MultClfs import *
#==================
# other vars
#==================
tts_split_sl = 'sl'
OutFile_suffix = 'sl'
#==================
# Specify outdir
#==================
outdir_ml = outdir + 'ml/tts_sl/'
print('\nOutput directory:', outdir_ml)
#outFile_wf = outdir_ml + gene.lower() + '_baselineC_' + OutFile_suffix + '.csv'
outFile_wf = outdir_ml + gene.lower() + '_baselineC_noOR' + OutFile_suffix + '.csv'
#%% Running models ############################################################
print('\n#####################################################################\n'
, '\nStarting--> Running ML analysis: Baseline modes (No FS)'
, '\nGene name:', gene
, '\nDrug name:', drug
, '\n#####################################################################\n')
paramD = {
'baseline_paramD': { 'input_df' : X
, 'target' : y
, 'var_type' : 'mixed'
, 'resampling_type': 'none'}
, 'smnc_paramD': { 'input_df' : X_smnc
, 'target' : y_smnc
, 'var_type' : 'mixed'
, 'resampling_type' : 'smnc'}
, 'ros_paramD': { 'input_df' : X_ros
, 'target' : y_ros
, 'var_type' : 'mixed'
, 'resampling_type' : 'ros'}
, 'rus_paramD' : { 'input_df' : X_rus
, 'target' : y_rus
, 'var_type' : 'mixed'
, 'resampling_type' : 'rus'}
, 'rouC_paramD' : { 'input_df' : X_rouC
, 'target' : y_rouC
, 'var_type' : 'mixed'
, 'resampling_type' : 'rouC'}
}
##==============================================================================
## Dict with no CV BT formatted df
## mmD = {}
## for k, v in paramD.items():
## # print(mmD[k])
## scores_slD = MultModelsCl(**paramD[k]
## , tts_split_type = tts_split_sl
## , skf_cv = skf_cv
## , blind_test_df = X_bts
## , blind_test_target = y_bts
## , add_cm = True
## , add_yn = True
## , return_formatted_output = False)
## mmD[k] = scores_slD
##==============================================================================
## Initial run to get the dict of dicts for each sampling type containing CV, BT and metadata DFs
mmDD = {}
for k, v in paramD.items():
scores_slD = MultModelsCl(**paramD[k]
, tts_split_type = tts_split_sl
, skf_cv = skf_cv
, blind_test_df = X_bts
, blind_test_target = y_bts
, add_cm = True
, add_yn = True
, return_formatted_output = True)
mmDD[k] = scores_slD
# Extracting the dfs from within the dict and concatenating to output as one df
for k, v in mmDD.items():
out_wf_sl = pd.concat(mmDD, ignore_index = True)
out_wf_slf = out_wf_sl.sort_values(by = ['resampling', 'source_data', 'MCC'], ascending = [True, True, False], inplace = False)
print('\n######################################################################'
, '\nEnd--> Successfully generated output DF for Multiple classifiers (baseline models)'
, '\nGene:', gene.lower()
, '\nDrug:', drug
, '\noutput file:', outFile_wf
, '\nDim of output:', out_wf_slf.shape
, '\n######################################################################')
###############################################################################
#====================
# Write output file
#====================
out_wf_slf.to_csv(outFile_wf, index = False)
print('\nFile successfully written:', outFile_wf)
###############################################################################

177
scripts/ml/running_ml_scripts.txt
View file

@ -1,158 +1,83 @@
########################################################################
#70/30
# 70/30 [WITHOUT OR]
########################################################################
=-----------------------------------=
# All features including AA index
# [WITH OR]
=-----------------------------------=
time ./run_7030.py -g pncA -d pyrazinamide 2>&1 | tee log_pnca_7030.txt #d
time ./run_7030.py -g embB -d ethambutol 2>&1 | tee log_embb_7030.txt
time ./run_7030.py -g katG -d isoniazid 2>&1 | tee log_katg_7030.txt
time ./run_7030.py -g rpoB -d rifampicin 2>&1 | tee log_rpob_7030.txt
time ./run_7030.py -g gid -d streptomycin 2>&1 | tee log_gid_7030.txt
time ./run_7030.py -g alr -d cycloserine 2>&1 | tee log_alr_7030.txt
# alr: # ERROR, as expected, too few values!
# gid: problems
=-----------------------------------=
# All features including AA index
# [WITHOUT OR] **DONE
# actual data
#------------------------------------=
time ./run_7030.py -g pncA -d pyrazinamide 2>&1 | tee log_pnca_7030_noOR.txt
time ./run_7030.py -g embB -d ethambutol 2>&1 | tee log_embb_7030_noOR.txt
time ./run_7030.py -g katG -d isoniazid 2>&1 | tee log_katg_7030_noOR.txt
time ./run_7030.py -g rpoB -d rifampicin 2>&1 | tee log_rpob_7030_noOR.txt
time ./run_7030.py -g gid -d streptomycin 2>&1 | tee log_gid_7030_noOR.txt
time ./run_7030.py -g alr -d cycloserine 2>&1 | tee log_alr_7030_noOR.txt
########################################################################
# 80/20
########################################################################
time ./run_7030.py -g pncA -d pyrazinamide 2>&1 | tee log_pnca_7030_.txt
time ./run_7030.py -g embB -d ethambutol 2>&1 | tee log_embb_7030_.txt
time ./run_7030.py -g katG -d isoniazid 2>&1 | tee log_katg_7030_.txt
time ./run_7030.py -g rpoB -d rifampicin 2>&1 | tee log_rpob_7030_.txt
time ./run_7030.py -g gid -d streptomycin 2>&1 | tee log_gid_7030_.txt
time ./run_7030.py -g alr -d cycloserine 2>&1 | tee log_alr_7030_.txt
=-----------------------------------=
# All features including AA index
# [WITH OR]
=-----------------------------------=
time ./run_8020.py -g pncA -d pyrazinamide 2>&1 | tee log_pnca_8020.txt
time ./run_8020.py -g embB -d ethambutol 2>&1 | tee log_embb_8020.txt
time ./run_8020.py -g katG -d isoniazid 2>&1 | tee log_katg_8020.txt
time ./run_8020.py -g rpoB -d rifampicin 2>&1 | tee log_rpob_8020.txt
time ./run_8020.py -g gid -d streptomycin 2>&1 | tee log_gid_8020.txt
time ./run_8020.py -g alr -d cycloserine 2>&1 | tee log_alr_8020.txt
=-----------------------------------=
# All features including AA index
# [WITHOUT OR] **DONE
real 0m1.099s
user 0m1.308s
sys 0m1.474s
=-----------------------------------=
time ./run_8020.py -g pncA -d pyrazinamide 2>&1 | tee log_pnca_8020_noOR.txt
time ./run_8020.py -g embB -d ethambutol 2>&1 | tee log_embb_8020_noOR.txt
time ./run_8020.py -g katG -d isoniazid 2>&1 | tee log_katg_8020_noOR.txt
time ./run_8020.py -g rpoB -d rifampicin 2>&1 | tee log_rpob_8020_noOR.txt
time ./run_8020.py -g gid -d streptomycin 2>&1 | tee log_gid_8020_noOR.txt
time ./run_8020.py -g alr -d cycloserine 2>&1 | tee log_alr_8020_noOR.txt
########################################################################
# SL
########################################################################
=-----------------------------------=
# All features including AA index
=-----------------------------------=
=-----------------------------------=
# All features including AA index
# [WITHOUT OR]
=-----------------------------------=
time ./run_sl.py -g pncA -d pyrazinamide 2>&1 | tee log_pnca_sl_noOR.txt
time ./run_sl.py -g embB -d ethambutol 2>&1 | tee log_embb_sl_noOR.txt
time ./run_sl.py -g katG -d isoniazid 2>&1 | tee log_katg_sl_noOR.txt
time ./run_sl.py -g rpoB -d rifampicin 2>&1 | tee log_rpob_sl_noOR.txt
time ./run_sl.py -g gid -d streptomycin 2>&1 | tee log_gid_sl_noOR.txt
time ./run_sl.py -g alr -d cycloserine 2>&1 | tee log_alr_sl_noOR.txt
########################################################################
########################################################################
########################################################################
###################### COMPLETE DATA ##############################
########################################################################
########################################################################
########################################################################
#70/30
########################################################################
=-----------------------------------=
# All features including AA index
# [WITHOUT OR]
# COMPLETE data
#------------------------------------=
time ./run_cd_7030.py -g pncA -d pyrazinamide 2>&1 | tee log_pnca_cd_7030_noOR.txt
time ./run_cd_7030.py -g embB -d ethambutol 2>&1 | tee log_embb_cd_7030_noOR.txt
time ./run_cd_7030.py -g katG -d isoniazid 2>&1 | tee log_katg_cd_7030_noOR.txt
time ./run_cd_7030.py -g rpoB -d rifampicin 2>&1 | tee log_rpob_cd_7030_noOR.txt
time ./run_cd_7030.py -g gid -d streptomycin 2>&1 | tee log_gid_cd_7030_noOR.txt
time ./run_cd_7030.py -g alr -d cycloserine 2>&1 | tee log_alr_cd_7030_noOR.txt
time ./run_cd_7030.py -g pncA -d pyrazinamide 2>&1 | tee log_pnca_cd_7030_.txt
time ./run_cd_7030.py -g embB -d ethambutol 2>&1 | tee log_embb_cd_7030_.txt
time ./run_cd_7030.py -g katG -d isoniazid 2>&1 | tee log_katg_cd_7030_.txt
time ./run_cd_7030.py -g rpoB -d rifampicin 2>&1 | tee log_rpob_cd_7030_.txt
time ./run_cd_7030.py -g gid -d streptomycin 2>&1 | tee log_gid_cd_7030_.txt
time ./run_cd_7030.py -g alr -d cycloserine 2>&1 | tee log_alr_cd_7030_.txt
########################################################################
# 80/20
# 80/20 [WITHOUT OR]
########################################################################
=-----------------------------------=
# actual data
#------------------------------------=
time ./run_8020.py -g pncA -d pyrazinamide 2>&1 | tee log_pnca_8020_.txt
time ./run_8020.py -g embB -d ethambutol 2>&1 | tee log_embb_8020_.txt
time ./run_8020.py -g katG -d isoniazid 2>&1 | tee log_katg_8020_.txt
time ./run_8020.py -g rpoB -d rifampicin 2>&1 | tee log_rpob_8020_.txt
time ./run_8020.py -g gid -d streptomycin 2>&1 | tee log_gid_8020_.txt
time ./run_8020.py -g alr -d cycloserine 2>&1 | tee log_alr_8020_.txt
=-----------------------------------=
# All features including AA index
# [WITHOUT OR]
# COMPLETE data
#------------------------------------=
time ./run_cd_8020.py -g pncA -d pyrazinamide 2>&1 | tee log_pnca_cd_8020_noOR.txt
time ./run_cd_8020.py -g embB -d ethambutol 2>&1 | tee log_embb_cd_8020_noOR.txt
time ./run_cd_8020.py -g katG -d isoniazid 2>&1 | tee log_katg_cd_8020_noOR.txt
time ./run_cd_8020.py -g rpoB -d rifampicin 2>&1 | tee log_rpob_cd_8020_noOR.txt
time ./run_cd_8020.py -g gid -d streptomycin 2>&1 | tee log_gid_cd_8020_noOR.txt
time ./run_cd_8020.py -g alr -d cycloserine 2>&1 | tee log_alr_cd_8020_noOR.txt
time ./run_cd_8020.py -g pncA -d pyrazinamide 2>&1 | tee log_pnca_cd_8020_.txt
time ./run_cd_8020.py -g embB -d ethambutol 2>&1 | tee log_embb_cd_8020_.txt
time ./run_cd_8020.py -g katG -d isoniazid 2>&1 | tee log_katg_cd_8020_.txt
time ./run_cd_8020.py -g rpoB -d rifampicin 2>&1 | tee log_rpob_cd_8020_.txt
time ./run_cd_8020.py -g gid -d streptomycin 2>&1 | tee log_gid_cd_8020_.txt
time ./run_cd_8020.py -g alr -d cycloserine 2>&1 | tee log_alr_cd_8020_.txt
########################################################################
# SL
# SL [WITHOUT OR]
########################################################################
=-----------------------------------=
# All features including AA index
# [WITHOUT OR]
# actual data
#------------------------------------=
time ./run_cd_sl.py -g pncA -d pyrazinamide 2>&1 | tee log_pnca_cd_sl_noOR.txt
time ./run_cd_sl.py -g embB -d ethambutol 2>&1 | tee log_embb_cd_sl_noOR.txt
time ./run_cd_sl.py -g katG -d isoniazid 2>&1 | tee log_katg_cd_sl_noOR.txt
time ./run_cd_sl.py -g rpoB -d rifampicin 2>&1 | tee log_rpob_cd_sl_noOR.txt
time ./run_cd_sl.py -g gid -d streptomycin 2>&1 | tee log_gid_cd_sl_noOR.txt
time ./run_cd_sl.py -g alr -d cycloserine 2>&1 | tee log_alr_cd_sl_noOR.txt
time ./run_sl.py -g pncA -d pyrazinamide 2>&1 | tee log_pnca_sl_.txt
time ./run_sl.py -g embB -d ethambutol 2>&1 | tee log_embb_sl_.txt
time ./run_sl.py -g katG -d isoniazid 2>&1 | tee log_katg_sl_.txt
time ./run_sl.py -g rpoB -d rifampicin 2>&1 | tee log_rpob_sl_.txt
time ./run_sl.py -g gid -d streptomycin 2>&1 | tee log_gid_sl_.txt
time ./run_sl.py -g alr -d cycloserine 2>&1 | tee log_alr_sl_.txt
=-----------------------------------=
# COMPLETE data
#------------------------------------=
time ./run_cd_sl.py -g pncA -d pyrazinamide 2>&1 | tee log_pnca_cd_sl_.txt
time ./run_cd_sl.py -g embB -d ethambutol 2>&1 | tee log_embb_cd_sl_.txt
time ./run_cd_sl.py -g katG -d isoniazid 2>&1 | tee log_katg_cd_sl_.txt
time ./run_cd_sl.py -g rpoB -d rifampicin 2>&1 | tee log_rpob_cd_sl_.txt
time ./run_cd_sl.py -g gid -d streptomycin 2>&1 | tee log_gid_cd_sl_.txt
time ./run_cd_sl.py -g alr -d cycloserine 2>&1 | tee log_alr_cd_sl_.txt
########################################################################
@ -167,4 +92,4 @@ time ./run_cd_sl.py -g alr -d cycloserine 2>&1 | tee log_alr_cd_sl_noOR.txt
time ./run_FS.py -g pncA -d pyrazinamide 2>&1 | tee log_FS_pnca_7030.txt
time ./run_FS_7030.py -g pncA -d pyrazinamide 2>&1 | tee log_FS_pnca_7030_noOR.txt
time ./run_FS_7030.py -g pncA -d pyrazinamide 2>&1 | tee log_FS_pnca_7030_.txt

116
scripts/ml/scrMult_CALL.py
View file

@ -1,116 +0,0 @@
fs_test = RFECV(DecisionTreeClassifier(**rs)
, cv = StratifiedKFold(n_splits = 10, shuffle = True,**rs)
, scoring = 'matthews_corrcoef')
models = [('Logistic Regression' , LogisticRegression(**rs) )]
#, ('Logistic RegressionCV' , LogisticRegressionCV(**rs) )]
for m in models:
print(m)
print('\n================================================================\n')
index = 1
for model_name, model_fn in models:
print('\nRunning classifier:', index
, '\nModel_name:' , model_name
, '\nModel func:' , model_fn)
#, '\nList of models:', models)
index = index+1
fs2 = RFECV(model_fn
, cv = skf_cv
, scoring = 'matthews_corrcoef')
from sklearn.datasets import make_friedman1
from sklearn.datasets import load_iris
X_eg, y_eg = load_iris(return_X_y=True)
#X_eg, y_eg = make_friedman1(n_samples=50, n_features=10, random_state=0)
fs2.fit(X_eg,y_eg)
fs2.support_
fs2.ranking_
###############################################################################
# LR
a_fs = fsgs(input_df = X
, target = y
#, param_gridLd = [{'fs__min_features_to_select' : []}]
, blind_test_df = X_bts
, blind_test_target = y_bts
#, estimator = RandomForestClassifier(**rs, **njobs, bootstrap = True, oob_score = True)
, estimator = LogisticRegression(**rs)
, use_fs = False # set True to use DT as a RFECV estimator
, var_type = 'mixed')
a_fs.keys()
a_fsDF = pd.DataFrame(a_fs.items()) # LR
a_fsDF2 = pd.DataFrame(a_fs2.items()) # use_FS= True
a_fsDF3 = pd.DataFrame(a_fs3.items()) # RF
# this one
a_fs0 = fsgs(input_df = X
, target = y
, param_gridLd = [{'fs__min_features_to_select' : [1]}]
, blind_test_df = X_bts
, blind_test_target = y_bts
, estimator = LogisticRegression(**rs)
, use_fs = False # uses estimator as the RFECV parameter for fs. Set to TRUE if you want to supply custom_fs as shown below
, custom_fs = RFECV(DecisionTreeClassifier(**rs) , cv = skf_cv, scoring = 'matthews_corrcoef')
, cv_method = skf_cv
, var_type = 'mixed'
)
###############################################
##############################################################################
# my function CALL
#import fsgs from UQ_FS_fn
# RFECV by default uses the estimator provided, custom option to provide fs model using use_fs and
a_fs = fsgs(input_df = X
, target = y
, param_gridLd = [{'fs__min_features_to_select' : [1]}]
, blind_test_df = X_bts
, blind_test_target = y_bts
, estimator = LogisticRegression(**rs)
#, use_fs = False # uses estimator as the RFECV parameter for fs. Set to TRUE if you want to supply custom_fs as shown below
, use_fs = True, custom_fs = RFECV(DecisionTreeClassifier(**rs) , cv = skf_cv, scoring = 'matthews_corrcoef')
, cv_method = skf_cv
, var_type = 'mixed'
)
a_fs.keys()
a_fs2.keys()
a_fs3.keys()
a_fsDF = pd.DataFrame(a_fs.items()) # LR
a_fsDF.columns = ['parameter', 'param_value']
a_fs2DF2 = pd.DataFrame(a_fs2.items()) # use_FS= True
a_fs2DF2.columns = ['parameter', 'param_value']
a_fsDF3 = pd.DataFrame(a_fs3.items()) # RF
##############
a_mask = a_fs['fs_res_array']
a_fsDF.loc[a_fsDF['parameter'] == 'fs_res_array']
mod_selF = a_fs2DF2.loc[a_fsDF['parameter'] == 'sel_features_names']; mod_selF
mod_selFT = mod_selF.T
# subset keys
#keys_to_extract = ['model_name', 'fs_method', 'sel_features_names', 'all_feature_names', 'fs_res_array']
keys_to_extract = ['fs_method', 'sel_features_names']
a_subset = {key: a_fs2[key] for key in keys_to_extract}
a_subsetDF = pd.DataFrame(a_subset); a_subsetDF
mod_fs_method = a_fs2['fs_method']
fs_name = re.search('estimator=(\w+)',mod_fs_method)
fs_namefN = fs_namef.group(1)
print('\nFS method:', fs_namefN)
fsDF = a_subsetDF[['sel_features_names']];fsDF
fsDF.columns = [fs_namefN+'_FS']
fsDF.columns; fsDF
###############################

117
scripts/ml/test_MultClfs.py
View file

@ -1,117 +0,0 @@
#!/usr/bin/env python3
# -*- coding: utf-8 -*-
"""
Created on Fri Jun 24 11:07:05 2022
@author: tanu
"""
import re
import argparse
import os, sys
###############################################################################
# gene = 'pncA'
# drug = 'pyrazinamide'
#total_mtblineage_uc = 8
# #%% command line args: case sensitive
# arg_parser = argparse.ArgumentParser()
# arg_parser.add_argument('-d', '--drug', help = 'drug name', default = '')
# arg_parser.add_argument('-g', '--gene', help = 'gene name', default = '')
# args = arg_parser.parse_args()
# drug = args.drug
# gene = args.gene
###############################################################################
homedir = os.path.expanduser("~")
sys.path.append(homedir + '/git/LSHTM_analysis/scripts/ml')
###############################################################################
#==================
# Import data
#==================
from ml_data_7030 import *
setvars(gene,drug)
from ml_data_7030 import *
# from YC run_all_ML: run locally
#from UQ_yc_RunAllClfs import run_all_ML
#====================
# Import ML functions
#====================
from MultClfs import *
#==================
# other vars
#==================
tts_split_7030 = '70_30'
OutFile_suffix = '7030'
#==================
# Specify outdir
#==================
outdir_ml = outdir + 'ml/tts_7030/'
print('\nOutput directory:', outdir_ml)
#outFile_wf = outdir_ml + gene.lower() + '_baselineC_' + OutFile_suffix + '.csv'
#outFile_lf = outdir_ml + gene.lower() + '_baselineC_ext_' + OutFile_suffix + '.csv'
###############################################################################
print('\n#####################################################################\n'
, '\nRunning ML analysis: Multiple models'
, '\nGene name:', gene
, '\nDrug name:', drug)
###############################################################################
#%% Test MultModelsCL WITHOUT returning formatted output
#================
# MultModelsCl: without formatted output
#================
mmD = MultModelsCl_noBT(input_df = X_smnc
, target = y_smnc
, var_type = 'mixed'
, tts_split_type = tts_split_7030
, resampling_type = 'smnc'
, skf_cv = skf_cv
, blind_test_df = X_bts
, blind_test_target = y_bts
, add_cm = True
, add_yn = True
, run_blind_test = True
, return_formatted_output = False)
#================
# MultModelsCl: WITH formatted output
#================
mmDF3 = MultModelsCl_noBT(input_df = X_smnc
, target = y_smnc
, var_type = 'mixed'
, tts_split_type = tts_split_7030
, resampling_type = 'smnc'
, skf_cv = skf_cv
, blind_test_df = X_bts
, blind_test_target = y_bts
, add_cm = True
, add_yn = True
, run_blind_test = True
, return_formatted_output= True )
mmDF9= MultModelsCl_noBT(input_df = X
, target = y
, var_type = 'mixed'
, tts_split_type = tts_split_7030
, resampling_type = 'none'
, skf_cv = None
, blind_test_df = X_bts
, blind_test_target = y_bts
, add_cm = True
, add_yn = True
, run_blind_test = True
, return_formatted_output= True )
#=================
# test function
#=================
# output from function call
ProcessMultModelsCl(mmD)
ProcessMultModelsCl(testD)