def _organize_variants(samples, batch_id):
"""Retrieve variant calls for all samples, merging batched samples into single VCF.
"""
bam_files = set([])
caller_names = [x["variantcaller"] for x in samples[0]["variants"]]
calls = collections.defaultdict(list)
for data in samples:
if "work_bam" in data:
bam_files.add(data["work_bam"])
for vrn in data["variants"]:
# for somatic ensemble, discard normal samples and filtered
# variants from vcfs
vrn_file = vrn["vrn_file"]
if data.get("metadata", False) and data["metadata"].get("phenotype", "normal").lower().startswith("tumor"):
vrn_file_temp = vrn_file.replace(".vcf", "_tumorOnly_noFilteredCalls.vcf") if ".vcf" in vrn_file else vrn_file_temp + "_tumorOnly_noFilteredCalls.vcf.gz"
# Select tumor sample and keep only PASS and . calls
vrn_file = vcfutils.select_sample(in_file=vrn_file, sample=data["name"][1],
out_file=vrn_file_temp,
config=data["config"], filters="PASS,.")
calls[vrn["variantcaller"]].append(vrn_file)
data = samples[0]
vrn_files = []
for caller in caller_names:
fnames = calls[caller]
if len(fnames) == 1:
vrn_files.append(fnames[0])
else:
vrn_files.append(population.get_multisample_vcf(fnames, batch_id, caller, data))
return caller_names, vrn_files, list(bam_files)
def _organize_variants(samples, batch_id):
"""Retrieve variant calls for all samples, merging batched samples into single VCF.
"""
bam_files = set([])
caller_names = [x["variantcaller"] for x in samples[0]["variants"]]
calls = collections.defaultdict(list)
for data in samples:
if "work_bam" in data:
bam_files.add(data["work_bam"])
for vrn in data["variants"]:
# for somatic ensemble, discard normal samples and filtered
# variants from vcfs
vrn_file = vrn["vrn_file"]
if data.get("metadata", False) and data["metadata"].get(
"phenotype", "normal").lower().startswith("tumor"):
vrn_file_temp = vrn_file.replace(
".vcf", "_tumorOnly_noFilteredCalls.vcf"
) if ".vcf" in vrn_file else vrn_file_temp + "_tumorOnly_noFilteredCalls.vcf.gz"
# Select tumor sample and keep only PASS and . calls
vrn_file = vcfutils.select_sample(in_file=vrn_file,
sample=data["name"][1],
out_file=vrn_file_temp,
config=data["config"],
filters="PASS,.")
calls[vrn["variantcaller"]].append(vrn_file)
data = samples[0]
vrn_files = []
for caller in caller_names:
fnames = calls[caller]
if len(fnames) == 1:
vrn_files.append(fnames[0])
else:
vrn_files.append(
population.get_multisample_vcf(fnames, batch_id, caller, data))
return caller_names, vrn_files, list(bam_files)
def _organize_variants(samples, batch_id):
"""Retrieve variant calls for all samples, merging batched samples into single VCF.
"""
caller_names = [x["variantcaller"] for x in samples[0]["variants"]]
calls = collections.defaultdict(list)
for data in samples:
for vrn in data["variants"]:
calls[vrn["variantcaller"]].append(vrn["vrn_file"])
data = samples[0]
vrn_files = []
for caller in caller_names:
fnames = calls[caller]
if len(fnames) == 1:
vrn_files.append(fnames[0])
else:
vrn_files.append(population.get_multisample_vcf(fnames, batch_id, caller, data))
return caller_names, vrn_files
