plink_dupvar_prioritize.py

#!/usr/bin/env python

import argparse

# Before running this script, generate *.dupvar and *.frq.counts files:
# 1) plink --bfile <PLINK file> --list-duplicate-vars
# 2) plink --bfile <PLINK file> --freq counts

# After running this script, you can exclude duplicated variant IDs with the highest missigness rate:
# 1) plink --bfile <PLINK file> --exclude <output> --keep-allele-order --make-bed --out <new PLINK file>

argparser = argparse.ArgumentParser(description = 'Selects variants with highest missigness from the duplicates (*.dupvar file).')
argparser.add_argument('-d', '--dupvar', metavar = 'file', dest = 'in_dupvar', type = str, required = True, help = '*.dupvar file generated by PLINK.')
argparser.add_argument('-c', '--counts', metavar = 'file', dest = 'in_counts', type = str, required = True, help = '*.frq.counts file generated by PLINK.')
argparser.add_argument('-o', '--output', metavar = 'file', dest = 'out_filename', type = str, required = True, help = 'Output file name with variant IDs.')

if __name__ == '__main__':
    args = argparser.parse_args()

    missingness = dict()

    with open(args.in_counts, 'rt') as icounts:
        header = icounts.readline().split()
        for line in icounts:
            fields = dict(zip(header, line.split()))
            missingness[fields['SNP']] = int(fields['G0'])


    with open(args.in_dupvar, 'rt') as idupvar, open(args.out_filename, 'wt') as ofile:
        header = idupvar.readline().split()
        for line in idupvar:
            var_ids = line.strip().split('\t')[-1].split()
            var_ids_sorted = sorted([(var_id, missingness[var_id]) for var_id in var_ids], key = lambda x: x[1], reverse = True)
            for var_id in var_ids_sorted[:-1]:
                ofile.write(var_id[0] + '\n')