geinf/vcf_to_sfs.py

57 lines
1.5 KiB
Python
Executable File

#!/usr/bin/env python3
"""
Caution : At the moment for gzipped files only.
ARGS
--------
usage : vcf_to_sfs.py VCF.gz nb_indiv
"""
import gzip
import sys
# default folded SFS
folded = True
diploid = True
phased = False
# PARAM : Nb of indiv
n = int(sys.argv[2])
if diploid and not folded:
n *= 2
# initiate SFS_values with a zeros dict
SFS_values = dict.fromkeys(range(n),0)
with gzip.open(sys.argv[1], "rb") as inputgz:
line = inputgz.readline()
genotypes = []
while line:
# decode gzipped binary lines
line = line.decode('utf-8').strip()
# every snp line, not comment or header
if not line.startswith("##") and not line.startswith("#"):
FORMAT = line.split("\t")[8:9]
SAMPLES = line.split("\t")[9:]
snp_genotypes = []
allele_counts = {}
for sample in SAMPLES:
# for UNPHASED data
smpl_genotype = [int(a) for a in sample.split(':')[0].split('/') if a != '.']
nb_alleles = set(smpl_genotype)
snp_genotypes += smpl_genotype
# skip if all individuals have the same genotype
if len(set(snp_genotypes)) == 1:
line = inputgz.readline()
continue
for k in set(snp_genotypes):
allele_counts[snp_genotypes.count(k)] = k
if folded :
SFS_values[min(allele_counts.keys())-1] += 1
line = inputgz.readline()
print(SFS_values)