richard
/
geinf
1
0
Fork 0
Code Issues Pull Requests Releases Wiki Activity

rectify sfs comp

master
tforest 2021-10-19 10:43:58 +02:00
parent 0e25049a58
commit a7490aed15
1 changed files with 18 additions and 6 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

24
vcf_to_sfs.py
View File

@ -38,13 +38,25 @@ with gzip.open(sys.argv[1], "rb") as inputgz:
smpl_genotype = [int(a) for a in sample.split(':')[0].split('/') if a != '.'] smpl_genotype = [int(a) for a in sample.split(':')[0].split('/') if a != '.']
nb_alleles = set(smpl_genotype) nb_alleles = set(smpl_genotype)
snp_genotypes += smpl_genotype snp_genotypes += smpl_genotype
if len(set(snp_genotypes)) == 1:
line = inputgz.readline()
continue
#print(snp_genotypes)
for k in set(snp_genotypes): for k in set(snp_genotypes):
allele_counts[snp_genotypes.count(k)] = k allele_counts[snp_genotypes.count(k)] = k
if 7 in allele_counts.keys():
print(allele_counts)
#print(allele_counts)
if folded : if folded :
for count in allele_counts.keys(): #for count in allele_counts.keys():
if count <= len(snp_genotypes)/2 : # for count in allele_counts.keys():
SFS_values[count-1] += 1 # if count <= len(snp_genotypes)/2 :
else: # SFS_values[count-1] += 1
SFS_values[len(snp_genotypes)-count] += 1 # else:
# SFS_values[len(snp_genotypes)-count-1] += 1
SFS_values[min(allele_counts.keys())-1] += 1
line = inputgz.readline() line = inputgz.readline()
print(SFS_values) #print(SFS_values)
# Note : tout est doublé là