format_vcf_for_CharGer.py

#!/usr/bin/env python

#title   :format_vcf_for_CharGer.py
#author  :Fernanda Martins Rodrigues (@fernanda);
#date    :20190820

"""
This script takes as input a VEP annotated VCF file and removes INFO fields generated by variant caller pipeline, leaving only VEP annotation fields.
    This formatting step is necessary prior to running CharGer, so it takes AF from gnomaD into account, other than AF from variant calling pipeline.
    Input must be a VEP annotated VCF file (.vcf or .vcf.gz accepted)

        Usage:
                python format_vcf_for_CharGer.py [-h] -i <input VCF file> -O <output directory>

        Arguments:
                -i, --inputVCF:          input VCF file; gzip compressed
                -O, --outputDirectory:   directory to write output files to
                -h, --help:            prints usage documentation

"""

import sys
import argparse
import getopt
import gzip
import os

def argument_parser():
    # create parser
    parser = argparse.ArgumentParser(description=__doc__)
    # add arguments
    parser.add_argument("-i", "--inputVCF", required=True, help="input VEP annotated VCF file; .vcf or .vcf.gz accepted")
    parser.add_argument("-O", "--outputDirectory", required=True, help="directory to write output files to")

    args = vars(parser.parse_args())
    inputVCF = args["inputVCF"]
    outputDirectory = args["outputDirectory"]

    if outputDirectory[-1] != '/':
        outputDirectory = outputDirectory + '/'

    if not os.path.exists(outputDirectory):
        os.makedirs(outputDirectory)

    return inputVCF, outputDirectory



###############
## MAIN CODE ##
###############


def main():
    inputVCF, outputDirectory = argument_parser()
    if ".gz" in inputVCF:
        try:
            vcfF = gzip.open(inputVCF,"rt")
        except IOError:
            print("VCF file does not exist!")
    else:
        try:
            vcfF = open(inputVCF,"r")
        except IOError:
            print("VCF file does not exist!")

    outFile_suffix = ".infoFixed.vcf"
    inputFile_basename = inputVCF.split('/')[-1]
    if ".gz" in inputVCF:
        outFile = outputDirectory+inputFile_basename.replace(".vcf.gz",outFile_suffix)
    else:
        outFile = outputDirectory+inputFile_basename.replace(".vcf",outFile_suffix)
    
    outF = open(outFile, "w")

    # Start parsing input VCF file
    for line in vcfF:
        # print the info lines to output file
        if line.startswith('#'):
            outF.write(line)
        else:
            var = line.strip().split("\t")
            info_VEPonly = var[7].split(";")[-1]
            var[7]=info_VEPonly
            outF.write('\t'.join(var)+'\n')
    outF.close()

if __name__ == "__main__":
        main()

## END ##################################