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

adding clean files for rerrun 35k dataset

Browse source
This commit is contained in:
Tanushree Tunstall 2020-07-07 18:28:55 +01:00
parent 943513a338
commit a7f21cfb14
32 changed files with 157 additions and 44550 deletions
Download patch file Download diff file Expand all files Collapse all files

230
scripts/kd_df.py
View file

@ -1,230 +0,0 @@
#!/usr/bin/env python3
# -*- coding: utf-8 -*-
'''
Created on Tue Aug 6 12:56:03 2019
@author: tanu
'''
#=======================================================================
# Task: Hydrophobicity (Kd) values for amino acid sequence using the
# Kyt&-Doolittle.
# Same output as using the expasy server (link below)
# Input: fasta file
# Output: csv file with
# useful links
# https://biopython.org/DIST/docs/api/Bio.SeqUtils.ProtParamData-pysrc.html
# https://web.expasy.org/protscale/pscale/protscale_help.html
#=======================================================================
#%% load packages
import sys, os
import argparse
import pandas as pd
import numpy as np
from pylab import *
from Bio.SeqUtils import ProtParamData
from Bio.SeqUtils.ProtParam import ProteinAnalysis
from Bio import SeqIO
#from Bio.Alphabet.IUPAC import IUPACProtein
import pprint as pp
#=======================================================================
#%% specify homedir and curr dir
homedir = os.path.expanduser('~')
# set working dir
os.getcwd()
os.chdir(homedir + '/git/LSHTM_analysis/scripts')
os.getcwd()
#=======================================================================
#%% command line args
arg_parser = argparse.ArgumentParser()
arg_parser.add_argument('-d', '--drug', help='drug name', default = None)
arg_parser.add_argument('-g', '--gene', help='gene name', default = None)
#arg_parser.add_argument('-p', '--plot', help='show plot', action='store_true')
args = arg_parser.parse_args()
#=======================================================================
#%% variable assignment: input and output
#drug = 'pyrazinamide'
#gene = 'pncA'
drug = args.drug
gene = args.gene
#plot = args.plot
gene_match = gene + '_p.'
#==========
# data dir
#==========
datadir = homedir + '/' + 'git/Data'
#=======
# input
#=======
indir = datadir + '/' + drug + '/' + 'input'
in_filename = '3pl1.fasta.txt'
infile = indir + '/' + in_filename
print('Input filename:', in_filename
, '\nInput path:', indir
, '\n============================================================')
#=======
# output
#=======
outdir = datadir + '/' + drug + '/' + 'output'
out_filename = gene.lower() + '_kd.csv'
outfile = outdir + '/' + out_filename
print('Output filename:', out_filename
, '\nOutput path:', outdir
, '\n=============================================================')
#%% end of variable assignment for input and output files
#=======================================================================
#%% kd values from fasta file and output csv
def kd_to_csv(inputfasta, outputkdcsv, windowsize = 3):
"""
Calculate kd (hydropathy values) from input fasta file
@param inputfasta: fasta file
@type inputfasta: string
@param outputkdcsv: csv file with kd values
@type outfile: string
@param windowsize: windowsize to perform KD calcs on (Kyte&-Doolittle)
@type DSSP: numeric
@return: none, writes kd values df as csv
"""
#========================
# read input fasta file
#========================
fh = open(inputfasta)
for record in SeqIO.parse(fh, 'fasta'):
id = record.id
seq = record.seq
num_residues = len(seq)
fh.close()
sequence = str(seq)
X = ProteinAnalysis(sequence)
#===================
# calculate KD values: same as the expasy server
#===================
my_window = windowsize
offset = round((my_window/2)-0.5)
# edge weight is set to default (100%)
kd_values = (X.protein_scale(ProtParamData.kd , window = my_window))
# sanity checks
print('Sequence Length:', num_residues)
print('kd_values Length:',len(kd_values))
print('Window Length:', my_window)
print('Window Offset:', offset)
print('=================================================================')
print('Checking:len(kd values) is as expected for the given window size & offset...')
expected_length = num_residues - (my_window - offset)
if len(kd_values) == expected_length:
print('PASS: expected and actual length of kd values match')
else:
print('FAIL: length mismatch'
,'\nExpected length:', expected_length
,'\nActual length:', len(kd_values)
, '\n=========================================================')
#===================
# creating two dfs
#===================
# 1) aa sequence and 2) kd_values. Then reset index for each df
# which will allow easy merging of the two dfs.
# df1: df of aa seq with index reset to start from 1
# (reflective of the actual aa position in a sequence)
# Name column of wt as 'wild_type' to be the same name used
# in the file required for merging later.
dfSeq = pd.DataFrame({'wild_type_kd':list(sequence)})
dfSeq.index = np.arange(1, len(dfSeq) + 1) # python is not inclusive
# df2: df of kd_values with index reset to start from offset + 1 and
# subsequent matched length of the kd_values
dfVals = pd.DataFrame({'kd_values':kd_values})
dfVals.index = np.arange(offset + 1, len(dfVals) + 1 + offset)
# sanity checks
max(dfVals['kd_values'])
min(dfVals['kd_values'])
#===================
# concatenating dfs
#===================
# Merge the two on index
# (as these are now reflective of the aa position numbers): df1 and df2
# This will introduce NaN where there is missing values. In our case this
# will be 2 (first and last ones based on window size and offset)
kd_df = pd.concat([dfSeq, dfVals], axis = 1)
#============================
# renaming index to position
#============================
kd_df = kd_df.rename_axis('position')
kd_df.head
print('Checking: position col i.e. index should be numeric')
if kd_df.index.dtype == 'int64':
print('PASS: position col is numeric'
, '\ndtype is:', kd_df.index.dtype)
else:
print('FAIL: position col is not numeric'
, '\nConverting to numeric')
kd_df.index.astype('int64')
print('Checking dtype for after conversion:\n'
, '\ndtype is:', kd_df.index.dtype
, '\n=========================================================')
#===============
# writing file
#===============
print('Writing file:'
, '\nFilename:', outputkdcsv
# , '\nPath:', outdir
, '\nExpected no. of rows:', len(kd_df)
, '\nExpected no. of cols:', len(kd_df.columns)
, '\n=============================================================')
kd_df.to_csv(outputkdcsv, header = True, index = True)
#===============
# plot: optional!
#===============
# http://www.dalkescientific.com/writings/NBN/plotting.html
# FIXME: save fig
# extract just pdb if from 'id' to pass to title of plot
# foo = re.match(r'(^[0-9]{1}\w{3})', id).groups(1)
# if doplot:
plot(kd_values, linewidth = 1.0)
#axis(xmin = 1, xmax = num_residues)
xlabel('Residue Number')
ylabel('Hydrophobicity')
title('K&D Hydrophobicity for ' + id)
show()
#%% end of function
#=======================================================================
#%% call function
#kd_to_csv(infile, outfile, windowsize = 3)
#=======================================================================
def main():
print('Running hydropathy calcs with following params\n'
, in_filename
, '\noutfile:', out_filename)
kd_to_csv(infile, outfile, 3)
print('Finished writing file:'
, '\nFilename:', outfile
, '\n=============================================================')
if __name__ == '__main__':
main()
#%% end of script
#=======================================================================