LSHTM_analysis/scripts/combining_dfs.py

497 lines
19 KiB
Python
Executable file

#!/usr/bin/env python3
# -*- coding: utf-8 -*-
'''
Created on Tue Aug 6 12:56:03 2019
@author: tanu
'''
#=======================================================================
# Task: combining all dfs to a single one
# Input: 8 dfs
#1) <gene>.lower()'_complex_mcsm_norm.csv'
#2) <gene>.lower()_foldx.csv'
#3) <gene>.lower()_dssp.csv'
#4) <gene>.lower()_kd.csv'
#5) <gene>.lower()_rd.csv'
#6) 'ns' + <gene>.lower()_snp_info.csv'
#7) <gene>.lower()_af_or.csv'
#8) <gene>.lower() _af_or_kinship.csv
# combining order
#Merge1 = 1 + 2
#Merge2 = 3 + 4
#Merge3 = Merge2 + 5
#Merge4 = Merge1 + Merge3
#Merge5 = 6 + 7
#Merge6 = Merge5 + 8
#Merge7 = Merge4 + Merge6
# Output: single csv of all 8 dfs combined
# useful link
# https://stackoverflow.com/questions/23668427/pandas-three-way-joining-multiple-dataframes-on-columns
#=======================================================================
#%% load packages
import sys, os
import pandas as pd
from pandas import DataFrame
import numpy as np
#from varname import nameof
import argparse
#=======================================================================
#%% specify input and curr dir
homedir = os.path.expanduser('~')
# set working dir
os.getcwd()
os.chdir(homedir + '/git/LSHTM_analysis/scripts')
os.getcwd()
# FIXME: local imports
#from combining import combine_dfs_with_checks
from combining_FIXME import detect_common_cols
from reference_dict import oneletter_aa_dict # CHECK DIR STRUC THERE!
#=======================================================================
#%% command line args
arg_parser = argparse.ArgumentParser()
arg_parser.add_argument('-d', '--drug', help='drug name', default = 'pyrazinamide')
arg_parser.add_argument('-g', '--gene', help='gene name', default = 'pncA') # case sensitive
arg_parser.add_argument('--datadir', help = 'Data Directory. By default, it assmumes homedir + git/Data')
arg_parser.add_argument('-i', '--input_dir', help = 'Input dir containing pdb files. By default, it assmumes homedir + <drug> + input')
arg_parser.add_argument('-o', '--output_dir', help = 'Output dir for results. By default, it assmes homedir + <drug> + output')
arg_parser.add_argument('--debug', action ='store_true', help = 'Debug Mode')
args = arg_parser.parse_args()
#=======================================================================
#%% variable assignment: input and output
#drug = 'pyrazinamide'
#gene = 'pncA'
gene_match = gene + '_p.'
drug = args.drug
gene = args.gene
datadir = args.datadir
indir = args.input_dir
outdir = args.output_dir
#%%=======================================================================
#==============
# directories
#==============
if not datadir:
datadir = homedir + '/' + 'git/Data'
if not indir:
indir = datadir + '/' + drug + '/input'
if not outdir:
outdir = datadir + '/' + drug + '/output'
#=======
# input
#=======
in_filename_mcsm = gene.lower() + '_complex_mcsm_norm.csv'
in_filename_foldx = gene.lower() + '_foldx.csv'
in_filename_dssp = gene.lower() + '_dssp.csv'
in_filename_kd = gene.lower() + '_kd.csv'
in_filename_rd = gene.lower() + '_rd.csv'
in_filename_snpinfo = 'ns' + gene.lower() + '_snp_info_f.csv'
in_filename_afor = gene.lower() + '_af_or.csv'
in_filename_afor_kin = gene.lower() + '_af_or_kinship.csv'
infile_mcsm = outdir + '/' + in_filename_mcsm
infile_foldx = outdir + '/' + in_filename_foldx
infile_dssp = outdir + '/' + in_filename_dssp
infile_kd = outdir + '/' + in_filename_kd
infile_rd = outdir + '/' + in_filename_rd
infile_snpinfo = outdir + '/' + in_filename_snpinfo
infile_afor = outdir + '/' + in_filename_afor
infile_afor_kin = outdir + '/' + in_filename_afor_kin
print('\nInput path:', indir
, '\nOutput path:', outdir, '\n'
, '\nInput filename mcsm:', infile_mcsm
, '\nInput filename foldx:', infile_foldx, '\n'
, '\nInput filename dssp:', infile_dssp
, '\nInput filename kd:', infile_kd
, '\nInput filename rd', infile_rd , '\n'
, '\nInput filename snp info:', infile_snpinfo, '\n'
, '\nInput filename af or:', infile_afor
, '\nInput filename afor kinship:', infile_afor_kin
, '\n============================================================')
#=======
# output
#=======
out_filename_comb = gene.lower() + '_all_params.csv'
outfile_comb = outdir + '/' + out_filename_comb
print('Output filename:', outfile_comb
, '\n===================================================================')
o_join = 'outer'
l_join = 'left'
r_join = 'right'
i_join = 'inner'
# end of variable assignment for input and output files
#%%============================================================================
print('==================================='
, '\nFirst merge: mcsm + foldx'
, '\n===================================')
mcsm_df = pd.read_csv(infile_mcsm, sep = ',')
#mcsm_df.columns = mcsm_df.columns.str.lower()
foldx_df = pd.read_csv(infile_foldx , sep = ',')
#mcsm_foldx_dfs = combine_dfs_with_checks(mcsm_df, foldx_df, my_join = o_join)
merging_cols_m1 = detect_common_cols(mcsm_df, foldx_df)
mcsm_foldx_dfs = pd.merge(mcsm_df, foldx_df, on = merging_cols_m1, how = o_join)
ncols_m1 = len(mcsm_foldx_dfs.columns)
print('\n\nResult of first merge:', mcsm_foldx_dfs.shape
, '\n===================================================================')
mcsm_foldx_dfs[merging_cols_m1].apply(len)
mcsm_foldx_dfs[merging_cols_m1].apply(len) == len(mcsm_foldx_dfs)
#%%============================================================================
print('==================================='
, '\nSecond merge: dssp + kd'
, '\n===================================')
dssp_df = pd.read_csv(infile_dssp, sep = ',')
kd_df = pd.read_csv(infile_kd, sep = ',')
rd_df = pd.read_csv(infile_rd, sep = ',')
#dssp_kd_dfs = combine_dfs_with_checks(dssp_df, kd_df, my_join = o_join)
merging_cols_m2 = detect_common_cols(dssp_df, kd_df)
dssp_kd_dfs = pd.merge(dssp_df, kd_df, on = merging_cols_m2, how = o_join)
print('\n\nResult of second merge:', dssp_kd_dfs.shape
, '\n===================================================================')
#%%============================================================================
print('==================================='
, '\nThird merge: second merge + rd_df'
, '\ndssp_kd_dfs + rd_df'
, '\n===================================')
#dssp_kd_rd_dfs = combine_dfs_with_checks(dssp_kd_dfs, rd_df, my_join = o_join)
merging_cols_m3 = detect_common_cols(dssp_df, kd_df)
dssp_kd_rd_dfs = pd.merge(dssp_kd_dfs, rd_df, on = merging_cols_m3, how = o_join)
ncols_m3 = len(dssp_kd_rd_dfs.columns)
print('\n\nResult of Third merge:', dssp_kd_rd_dfs.shape
, '\n===================================================================')
dssp_kd_rd_dfs[merging_cols_m3].apply(len)
dssp_kd_rd_dfs[merging_cols_m3].apply(len) == len(dssp_kd_rd_dfs)
#%%============================================================================
print('======================================='
, '\nFourth merge: First merge + Third merge'
, '\nmcsm_foldx_dfs + dssp_kd_rd_dfs'
, '\n=======================================')
#combined_df = combine_dfs_with_checks(mcsm_foldx_dfs, dssp_kd_rd_dfs, my_join = i_join)
merging_cols_m4 = detect_common_cols(mcsm_foldx_dfs, dssp_kd_rd_dfs)
combined_df = pd.merge(mcsm_foldx_dfs, dssp_kd_rd_dfs, on = merging_cols_m4, how = i_join)
combined_df_expected_cols = ncols_m1 + ncols_m3 - len(merging_cols_m4)
if len(combined_df) == len(mcsm_df) and len(combined_df.columns) == combined_df_expected_cols:
print('PASS: successfully combined 5 dfs'
, '\nNo. of rows combined_df:', len(combined_df)
, '\nNo. of cols combined_df:', len(combined_df.columns))
else:
sys.exit('FAIL: check individual df merges')
print('\nResult of Fourth merge:', combined_df.shape
, '\n===================================================================')
combined_df[merging_cols_m4].apply(len)
combined_df[merging_cols_m4].apply(len) == len(combined_df)
#%%============================================================================
# OR merges: TEDIOUSSSS!!!!
#%%
print('==================================='
, '\nFifth merge: afor_df + afor_kin_df'
, '\n===================================')
# OR combining
afor_df = pd.read_csv(infile_afor, sep = ',')
#afor_df.columns = afor_df.columns.str.lower()
afor_kin_df = pd.read_csv(infile_afor_kin, sep = ',')
afor_kin_df.columns = afor_kin_df.columns.str.lower()
merging_cols_m5 = detect_common_cols(afor_df, afor_kin_df)
print('Dim of afor_df:', afor_df.shape
, '\nDim of afor_kin_df:', afor_kin_df.shape)
# finding if ALL afor_kin_df muts are present in afor_df
# i.e all kinship muts should be PRESENT in mycalcs_present
if len(afor_kin_df[afor_kin_df['mutation'].isin(afor_df['mutation'])]) == afor_kin_df.shape[0]:
print('PASS: ALL or_kinship muts are present in my or list')
else:
nf_muts = len(afor_kin_df[~afor_kin_df['mutation'].isin(afor_df['mutation'])])
nf_muts_df = afor_kin_df[~afor_kin_df['mutation'].isin(afor_df['mutation'])]
print('FAIL:', nf_muts, 'muts present in afor_kin_df NOT present in my or list'
, '\nsee "nf_muts_df" created containing not found(nf) muts')
sys.exit()
# Now checking how many afor_df muts are NOT present in afor_kin_df
common_muts = len(afor_df[afor_df['mutation'].isin(afor_kin_df['mutation'])])
extra_muts_myor = afor_kin_df.shape[0] - common_muts
print('=========================================='
, '\nmy or calcs has', extra_muts_myor, 'extra mutations'
, '\n==========================================')
print('Expected cals for merging with outer_join...')
expected_rows = afor_df.shape[0] + extra_muts_myor
expected_cols = afor_df.shape[1] + afor_kin_df.shape[1] - len(merging_cols_m5)
ors_df = pd.merge(afor_df, afor_kin_df, on = merging_cols_m5, how = o_join)
if ors_df.shape[0] == expected_rows and ors_df.shape[1] == expected_cols:
print('PASS: OR dfs successfully combined! PHEWWWW!')
else:
print('FAIL: could not combine OR dfs'
, '\nCheck expected rows and cols calculation and join type')
print('Dim of merged ors_df:', ors_df.shape)
ors_df[merging_cols_m5].apply(len)
ors_df[merging_cols_m5].apply(len) == len(ors_df)
#%%============================================================================
# formatting ors_df
ors_df.columns
# Dropping unncessary columns: already removed in ealier preprocessing
#cols_to_drop = ['reference_allele', 'alternate_allele', 'symbol' ]
cols_to_drop = ['n_miss']
print('Dropping', len(cols_to_drop), 'columns:\n'
, cols_to_drop)
ors_df.drop(cols_to_drop, axis = 1, inplace = True)
print('Reordering', ors_df.shape[1], 'columns'
, '\n===============================================')
cols = ors_df.columns
column_order = ['mutation'
, 'mutationinformation'
, 'wild_type'
, 'position'
, 'mutant_type'
#, 'chr_num_allele' #old
, 'ref_allele'
, 'alt_allele'
, 'mut_info_f1'
, 'mut_info_f2'
, 'mut_type'
, 'gene_id'
#, 'gene_number' #old
, 'gene_name'
#, 'mut_region'
#, 'reference_allele'
#, 'alternate_allele'
, 'chromosome_number'
, 'af'
, 'af_kin'
, 'est_chisq'
, 'or_mychisq'
, 'or_fisher'
, 'or_logistic'
, 'or_kin'
, 'pval_chisq'
, 'pval_fisher'
, 'pval_logistic'
, 'pwald_kin'
, 'ci_low_fisher'
, 'ci_hi_fisher'
, 'ci_low_logistic'
, 'ci_hi_logistic'
, 'beta_logistic'
, 'beta_kin'
, 'se_logistic'
, 'se_kin'
, 'zval_logistic'
, 'logl_h1_kin'
, 'l_remle_kin'
#, 'wt_3let' # old
#, 'mt_3let' # old
#, 'symbol'
#, 'n_miss'
]
if ( (len(column_order) == ors_df.shape[1]) and (DataFrame(column_order).isin(ors_df.columns).all().all())):
print('PASS: Column order generated for all:', len(column_order), 'columns'
, '\nColumn names match, safe to reorder columns'
, '\nApplying column order to df...' )
ors_df_ordered = ors_df[column_order]
else:
print('FAIL: Mismatch in no. of cols to reorder'
, '\nNo. of cols in df to reorder:', ors_df.shape[1]
, '\nNo. of cols order generated for:', len(column_order))
sys.exit()
print('\nResult of Sixth merge:', ors_df_ordered.shape
, '\n===================================================================')
#%%============================================================================
print('==================================='
, '\nSixth merge: Fourth + Fifth merge'
, '\ncombined_df + ors_df_ordered'
, '\n===================================')
#combined_df_all = combine_dfs_with_checks(combined_df, ors_df_ordered, my_join = i_join)
merging_cols_m6 = detect_common_cols(combined_df, ors_df_ordered)
print('Dim of df1:', combined_df.shape
, '\nDim of df2:', ors_df_ordered.shape
, '\nNo. of merging_cols:', len(merging_cols_m6))
print('Checking mutations in the two dfs:'
, '\nmuts in df1 but NOT in df2:'
, combined_df['mutationinformation'].isin(ors_df_ordered['mutationinformation']).sum()
, '\nmuts in df2 but NOT in df1:'
, ors_df_ordered['mutationinformation'].isin(combined_df['mutationinformation']).sum())
#print('\nNo. of common muts:', np.intersect1d(combined_df['mutationinformation'], ors_df_ordered['mutationinformation']) )
combined_df_all = pd.merge(combined_df, ors_df, on = merging_cols_m6, how = l_join)
combined_df_all.shape
# populate mut_info_f1
combined_df_all['mut_info_f1'].isna().sum()
combined_df_all['mut_info_f1'] = combined_df_all['position'].astype(str) + combined_df_all['wild_type'] + '>' + combined_df_all['position'].astype(str) + combined_df_all['mutant_type']
combined_df_all['mut_info_f1'].isna().sum()
# populate mut_type
combined_df_all['mut_type'].isna().sum()
#mut_type_word = combined_df_all['mut_type'].value_counts()
mut_type_word = 'missense' # FIXME, should be derived
combined_df_all['mut_type'].fillna(mut_type_word, inplace = True)
combined_df_all['mut_type'].isna().sum()
# populate gene_id
combined_df_all['gene_id'].isna().sum()
#gene_id_word = combined_df_all['gene_id'].value_counts()
gene_id_word = 'Rv2043c' # FIXME, should be derived
combined_df_all['gene_id'].fillna(gene_id_word, inplace = True)
combined_df_all['gene_id'].isna().sum()
# populate gene_name
combined_df_all['gene_name'].isna().sum()
combined_df_all['gene_name'].value_counts()
combined_df_all['gene_name'].fillna(gene, inplace = True)
combined_df_all['gene_name'].isna().sum()
# FIXME: DIM
# only with left join!
outdf_expected_rows = len(combined_df)
outdf_expected_cols = len(combined_df.columns) + len(ors_df_ordered.columns) - len(merging_cols_m6)
#if combined_df_all.shape[1] == outdf_expected_cols and combined_df_all.shape[0] == outdf_expected_rows:
if combined_df_all.shape[1] == outdf_expected_cols and combined_df_all['mutationinformation'].nunique() == outdf_expected_rows:
print('PASS: Df dimension match'
, '\nDim of combined_df_all with join type:', l_join
, '\n', combined_df_all.shape
, '\n===============================================================')
else:
print('FAIL: Df dimension mismatch'
, 'Cannot generate expected dim. See details of merge performed'
, '\ndf1 dim:', combined_df.shape
, '\ndf2 dim:', ors_df.shape
, '\nGot:', combined_df_all.shape
, '\nmuts in df1 but NOT in df2:'
, combined_df['mutationinformation'].isin(ors_df['mutationinformation']).sum()
, '\nmuts in df2 but NOT in df1:'
, ors_df['mutationinformation'].isin(combined_df['mutationinformation']).sum())
sys.exit()
#%% IMPORTANT: check if mutation related info is all populated after this merge
# FIXME: should get fixed with JP's resolved dataset!?
check_nan = combined_df_all.isna().sum(axis = 0)
# relevant mut cols
check_mut_cols = merging_cols_m5 + merging_cols_m6
count_na_mut_cols = combined_df_all[check_mut_cols].isna().sum().reset_index().rename(columns = {'index': 'col_name', 0: 'na_count'})
if (count_na_mut_cols['na_count'].sum() > 0).any():
# FIXME: static override, generate 'mutation' from variable
na_muts_n = combined_df_all['mutation'].isna().sum()
baz = combined_df_all[combined_df_all['mutation'].isna()]
baz = baz[check_mut_cols]
print(na_muts_n, 'mutations have missing \'mutation\' info.'
, '\nFetching these from reference dict...')
lookup_dict = dict()
for k, v in oneletter_aa_dict.items():
lookup_dict[k] = v['three_letter_code_lower']
print(lookup_dict)
wt_3let = combined_df_all['wild_type'].map(lookup_dict).str.capitalize()
#print(wt_3let)
pos = combined_df_all['position'].astype(str)
#print(pos)
mt_3let = combined_df_all['mutant_type'].map(lookup_dict).str.capitalize()
#print(mt_3let)
baz['mutation'] = 'pnca_p.' + wt_3let + pos + mt_3let
print(combined_df_all['mutation'])
# populate mut_info_f2
combined_df_all['mut_info_f2'] = combined_df_all['mutation'].str.replace(gene_match.lower(), 'p.', regex = True)
#%% merge
#merging_cols_m7 = detect_common_cols(combined_df_all, baz)
baz2 = baz[['mutationinformation', 'mut_info_f2']]
baz2 = baz2.drop_duplicates()
merging_cols_m7 = detect_common_cols(combined_df_all, baz2)
combined_df_all2 = pd.merge(combined_df_all, baz2
#, on = merging_cols_m7
, on = 'mutationinformation'
, how = o_join)
#%%============================================================================
output_cols = combined_df_all.columns
print('Output cols:', output_cols)
#%%============================================================================
# write csv
print('Writing file: combined output of all params needed for plotting and ML')
combined_df_all.to_csv(outfile_comb, index = False)
print('\nFinished writing file:'
, '\nNo. of rows:', combined_df_all.shape[0]
, '\nNo. of cols:', combined_df_all.shape[1])
#=======================================================================
#%% incase you FIX the the function: combine_dfs_with_checks
#def main():
# print('Reading input files:')
#mcsm_df = pd.read_csv(infile_mcsm, sep = ',')
#mcsm_df.columns = mcsm_df.columns.str.lower()
#foldx_df = pd.read_csv(infile_foldx , sep = ',')
#dssp_df = pd.read_csv(infile_dssp, sep = ',')
#dssp_df.columns = dssp_df.columns.str.lower()
#kd_df = pd.read_csv(infile_kd, sep = ',')
#kd_df.columns = kd_df.columns.str.lower()
#rd_df = pd.read_csv(infile_kd, sep = ',')
#if __name__ == '__main__':
# main()
#=======================================================================
#%% end of script