def variantcall_sample(data, region=None, align_bams=None, out_file=None):
"""Parallel entry point for doing genotyping of a region of a sample.
"""
if out_file is None or not os.path.exists(out_file) or not os.path.lexists(
out_file):
utils.safe_makedir(os.path.dirname(out_file))
sam_ref = data["sam_ref"]
config = data["config"]
caller_fns = get_variantcallers()
caller_fn = caller_fns[config["algorithm"].get("variantcaller",
"gatk")]
if len(align_bams) == 1:
items = [data]
else:
items = multi.get_orig_items(data)
assert len(items) == len(align_bams)
call_file = "%s-raw%s" % utils.splitext_plus(out_file)
call_file = caller_fn(align_bams, items, sam_ref,
data["genome_resources"]["variation"], region,
call_file)
if data["config"]["algorithm"].get("phasing", False) == "gatk":
call_file = phasing.read_backed_phasing(call_file, align_bams,
sam_ref, region, config)
utils.symlink_plus(call_file, out_file)
if region:
data["region"] = region
data["vrn_file"] = out_file
return [data]
def variantcall_sample(data, region=None, align_bams=None, out_file=None):
"""Parallel entry point for doing genotyping of a region of a sample.
"""
if out_file is None or not os.path.exists(out_file) or not os.path.lexists(
out_file):
utils.safe_makedir(os.path.dirname(out_file))
ref_file = dd.get_ref_file(data)
config = data["config"]
caller_fns = get_variantcallers()
caller_fn = caller_fns[config["algorithm"].get("variantcaller")]
if len(align_bams) == 1:
items = [data]
else:
items = multi.get_orig_items(data)
assert len(items) == len(align_bams)
assoc_files = tz.get_in(("genome_resources", "variation"), data, {})
if not assoc_files: assoc_files = {}
for bam_file in align_bams:
bam.index(bam_file, data["config"], check_timestamp=False)
out_file = caller_fn(align_bams, items, ref_file, assoc_files, region,
out_file)
if region:
data["region"] = region
data["vrn_file"] = out_file
return [data]
def variantcall_sample(data, region=None, align_bams=None, out_file=None):
"""Parallel entry point for doing genotyping of a region of a sample.
"""
if out_file is None or not os.path.exists(out_file) or not os.path.lexists(out_file):
utils.safe_makedir(os.path.dirname(out_file))
sam_ref = data["sam_ref"]
config = data["config"]
caller_fns = get_variantcallers()
caller_fn = caller_fns[config["algorithm"].get("variantcaller", "gatk")]
if len(align_bams) == 1:
items = [data]
else:
items = multi.get_orig_items(data)
assert len(items) == len(align_bams)
assoc_files = tz.get_in(("genome_resources", "variation"), data, {})
if not assoc_files: assoc_files = {}
for bam_file in align_bams:
bam.index(bam_file, data["config"], check_timestamp=False)
do_phasing = data["config"]["algorithm"].get("phasing", False)
call_file = "%s-raw%s" % utils.splitext_plus(out_file) if do_phasing else out_file
call_file = caller_fn(align_bams, items, sam_ref, assoc_files, region, call_file)
if do_phasing == "gatk":
call_file = phasing.read_backed_phasing(call_file, align_bams, sam_ref, region, config)
utils.symlink_plus(call_file, out_file)
if region:
data["region"] = region
data["vrn_file"] = out_file
return [data]
def _get_orig_items(data):
"""Retrieve original items in a batch, handling CWL and standard cases.
"""
if isinstance(data, dict):
if dd.get_align_bam(data) and tz.get_in(["metadata", "batch"], data):
return vmulti.get_orig_items(data)
else:
return [data]
else:
return data
def _get_orig_items(data):
"""Retrieve original items in a batch, handling CWL and standard cases.
"""
if isinstance(data, dict):
if tz.get_in(["metadata", "batch"], data):
return vmulti.get_orig_items(data)
else:
return [data]
else:
return data
def _get_validate(data):
"""Retrieve items to validate, from single samples or from combined joint calls.
"""
if data.get("vrn_file") and "validate" in data["config"]["algorithm"]:
return data
elif "group_orig" in data:
for sub in multi.get_orig_items(data):
if "validate" in sub["config"]["algorithm"]:
return sub
return None
def _get_validate(data):
"""Retrieve items to validate, from single samples or from combined joint calls.
"""
if data.get("vrn_file") and "validate" in data["config"]["algorithm"]:
return data
elif "group_orig" in data:
for sub in multi.get_orig_items(data):
if "validate" in sub["config"]["algorithm"]:
sub_val = utils.deepish_copy(sub)
sub_val["vrn_file"] = data["vrn_file"]
return sub_val
return None
def _get_validate(data):
"""Retrieve items to validate, from single samples or from combined joint calls.
"""
if data.get("vrn_file") and tz.get_in(["config", "algorithm", "validate"], data):
return data
elif "group_orig" in data:
for sub in multi.get_orig_items(data):
if "validate" in sub["config"]["algorithm"]:
sub_val = utils.deepish_copy(sub)
sub_val["vrn_file"] = data["vrn_file"]
return sub_val
return None
def extract(data):
"""Extract germline calls for the given sample, if tumor/normal or prioritized.
"""
if vcfutils.get_paired_phenotype(data):
is_paired = dd.get_batches(data) and len(vmulti.get_orig_items(data)) > 1
if is_paired:
germline_vcf = _extract_germline(data["vrn_file"], data)
else:
germline_vcf = _remove_prioritization(data["vrn_file"], data)
germline_vcf = vcfutils.bgzip_and_index(germline_vcf, data["config"])
data["vrn_file_plus"] = {"germline": germline_vcf}
return data
def extract(data):
"""Extract germline calls for the given sample, if tumor/normal or prioritized.
"""
if vcfutils.get_paired_phenotype(data):
is_paired = dd.get_batches(data) and len(
vmulti.get_orig_items(data)) > 1
if is_paired:
germline_vcf = _extract_germline(data["vrn_file"], data)
else:
germline_vcf = _remove_prioritization(data["vrn_file"], data)
germline_vcf = vcfutils.bgzip_and_index(germline_vcf, data["config"])
data["vrn_file_plus"] = {"germline": germline_vcf}
return data
def _do_prioritize(data):
"""Determine if we should perform prioritization.
Currently done on tumor-only input samples.
"""
if vcfutils.get_paired_phenotype(data):
has_tumor = False
has_normal = False
orig_items = vmulti.get_orig_items(data) if tz.get_in(["metadata", "batch"], data) else [data]
for sub_data in orig_items:
if vcfutils.get_paired_phenotype(sub_data) == "tumor":
has_tumor = True
elif vcfutils.get_paired_phenotype(sub_data) == "normal":
has_normal = True
return has_tumor and not has_normal
def _do_prioritize(data):
"""Determine if we should perform prioritization.
Currently done on tumor-only input samples.
"""
if vcfutils.get_paired_phenotype(data):
has_tumor = False
has_normal = False
orig_items = vmulti.get_orig_items(data) if tz.get_in(
["metadata", "batch"], data) else [data]
for sub_data in orig_items:
if vcfutils.get_paired_phenotype(sub_data) == "tumor":
has_tumor = True
elif vcfutils.get_paired_phenotype(sub_data) == "normal":
has_normal = True
return has_tumor and not has_normal
