def _shared_gatk_call_prep(align_bams, items, ref_file, region, out_file, num_cores=1):
"""Shared preparation work for GATK variant calling.
"""
data = items[0]
config = data["config"]
broad_runner = broad.runner_from_config(config)
gatk_type = broad_runner.gatk_type()
for x in align_bams:
bam.index(x, config)
picard_runner = broad.runner_from_path("picard", config)
picard_runner.run_fn("picard_index_ref", ref_file)
params = ["-R", ref_file]
coverage_depth_min = tz.get_in(["algorithm", "coverage_depth_min"], config)
if coverage_depth_min and coverage_depth_min < 4:
confidence = "4.0"
params += ["--standard_min_confidence_threshold_for_calling", confidence]
for a in annotation.get_gatk_annotations(config):
params += ["--annotation", a]
for x in align_bams:
params += ["-I", x]
variant_regions = bedutils.population_variant_regions(items)
region = subset_variant_regions(variant_regions, region, out_file, items)
if region:
if gatk_type == "gatk4":
params += ["-L", bamprep.region_to_gatk(region), "--interval-set-rule", "INTERSECTION"]
else:
params += ["-L", bamprep.region_to_gatk(region), "--interval_set_rule", "INTERSECTION"]
params += standard_cl_params(items)
return broad_runner, params
def mutect2_caller(align_bams, items, ref_file, assoc_files,
region=None, out_file=None):
"""Call variation with GATK's MuTect2.
This requires the full non open-source version of GATK 3.5+.
"""
if out_file is None:
out_file = "%s-variants.vcf.gz" % utils.splitext_plus(align_bams[0])[0]
if not utils.file_exists(out_file):
_prep_inputs(align_bams, ref_file, items)
with file_transaction(items[0], out_file) as tx_out_file:
params = ["-T", "MuTect2",
"-R", ref_file,
"--annotation", "ClippingRankSumTest",
"--annotation", "DepthPerSampleHC"]
for a in annotation.get_gatk_annotations(items[0]["config"]):
params += ["--annotation", a]
paired = vcfutils.get_paired_bams(align_bams, items)
params += _add_tumor_params(paired)
params += _add_region_params(region, out_file, items)
params += _add_assoc_params(assoc_files)
params += ["-ploidy", str(ploidy.get_ploidy(items, region))]
resources = config_utils.get_resources("mutect2", items[0]["config"])
if "options" in resources:
params += [str(x) for x in resources.get("options", [])]
broad_runner = broad.runner_from_config(items[0]["config"])
assert LooseVersion(broad_runner.gatk_major_version()) >= LooseVersion("3.5"), \
"Require full version of GATK 3.5+ for mutect2 calling"
broad_runner.new_resources("mutect2")
gatk_cmd = " ".join(broad_runner.cl_gatk(params, os.path.dirname(tx_out_file)))
pp_cmd = _post_process_cl(paired)
cmd = "{gatk_cmd} | {pp_cmd} | bgzip -c > {tx_out_file}"
do.run(cmd.format(**locals()), "MuTect2")
out_file = vcfutils.bgzip_and_index(out_file, items[0]["config"])
return out_file
def _shared_gatk_call_prep(align_bams, items, ref_file, dbsnp, region,
out_file):
"""Shared preparation work for GATK variant calling.
"""
data = items[0]
config = data["config"]
broad_runner = broad.runner_from_path("picard", config)
broad_runner.run_fn("picard_index_ref", ref_file)
for x in align_bams:
bam.index(x, config)
params = ["-R", ref_file]
coverage_depth_min = tz.get_in(["algorithm", "coverage_depth_min"], config)
if coverage_depth_min and coverage_depth_min < 4:
confidence = "4.0"
params += [
"--standard_min_confidence_threshold_for_calling", confidence
]
for a in annotation.get_gatk_annotations(config):
params += ["--annotation", a]
for x in align_bams:
params += ["-I", x]
if dbsnp:
params += ["--dbsnp", dbsnp]
variant_regions = bedutils.population_variant_regions(items)
region = subset_variant_regions(variant_regions, region, out_file, items)
if region:
params += [
"-L",
bamprep.region_to_gatk(region), "--interval_set_rule",
"INTERSECTION"
]
params += standard_cl_params(items)
broad_runner = broad.runner_from_config(config)
return broad_runner, params
def _shared_gatk_call_prep(align_bams, items, ref_file, dbsnp, region, out_file):
"""Shared preparation work for GATK variant calling.
"""
data = items[0]
config = data["config"]
broad_runner = broad.runner_from_path("picard", config)
broad_runner.run_fn("picard_index_ref", ref_file)
for x in align_bams:
bam.index(x, config)
params = ["-R", ref_file]
coverage_depth_min = tz.get_in(["algorithm", "coverage_depth_min"], config)
if coverage_depth_min and coverage_depth_min < 4:
confidence = "4.0"
params += ["--standard_min_confidence_threshold_for_calling", confidence,
"--standard_min_confidence_threshold_for_emitting", confidence]
for a in annotation.get_gatk_annotations(config):
params += ["--annotation", a]
for x in align_bams:
params += ["-I", x]
if dbsnp:
params += ["--dbsnp", dbsnp]
variant_regions = tz.get_in(["algorithm", "variant_regions"], config)
region = subset_variant_regions(variant_regions, region, out_file, items)
if region:
params += ["-L", bamprep.region_to_gatk(region), "--interval_set_rule", "INTERSECTION"]
broad_runner = broad.runner_from_config(config)
return broad_runner, params
def _shared_gatk_call_prep(align_bams, ref_file, config, dbsnp, region, out_file):
"""Shared preparation work for GATK variant calling.
"""
broad_runner = broad.runner_from_config(config)
broad_runner.run_fn("picard_index_ref", ref_file)
for x in align_bams:
bam.index(x, config)
coverage_depth = config["algorithm"].get("coverage_depth", "high").lower()
variant_regions = config["algorithm"].get("variant_regions", None)
confidence = "4.0" if coverage_depth in ["low"] else "30.0"
region = subset_variant_regions(variant_regions, region, out_file)
params = ["-R", ref_file,
"--standard_min_confidence_threshold_for_calling", confidence,
"--standard_min_confidence_threshold_for_emitting", confidence,
"--downsample_to_coverage", "250",
"--downsampling_type", "BY_SAMPLE",
]
for a in annotation.get_gatk_annotations(config):
params += ["--annotation", a]
for x in align_bams:
params += ["-I", x]
if dbsnp:
params += ["--dbsnp", dbsnp]
if region:
params += ["-L", bamprep.region_to_gatk(region), "--interval_set_rule", "INTERSECTION"]
return broad_runner, params
def _shared_gatk_call_prep(align_bams, items, ref_file, dbsnp, region, out_file):
"""Shared preparation work for GATK variant calling.
"""
config = items[0]["config"]
broad_runner = broad.runner_from_config(config)
broad_runner.run_fn("picard_index_ref", ref_file)
for x in align_bams:
bam.index(x, config)
# GATK can only downsample to a minimum of 200
coverage_depth_max = max(200, utils.get_in(config, ("algorithm", "coverage_depth_max"), 2000))
coverage_depth_min = utils.get_in(config, ("algorithm", "coverage_depth_min"), 4)
variant_regions = config["algorithm"].get("variant_regions", None)
confidence = "4.0" if coverage_depth_min < 4 else "30.0"
region = subset_variant_regions(variant_regions, region, out_file, items)
params = ["-R", ref_file,
"--standard_min_confidence_threshold_for_calling", confidence,
"--standard_min_confidence_threshold_for_emitting", confidence,
"--downsample_to_coverage", str(coverage_depth_max),
"--downsampling_type", "BY_SAMPLE",
]
for a in annotation.get_gatk_annotations(config):
params += ["--annotation", a]
for x in align_bams:
params += ["-I", x]
if dbsnp:
params += ["--dbsnp", dbsnp]
if region:
params += ["-L", bamprep.region_to_gatk(region), "--interval_set_rule", "INTERSECTION"]
return broad_runner, params
def _shared_gatk_call_prep(align_bams, ref_file, config, dbsnp, region, out_file):
"""Shared preparation work for GATK variant calling.
"""
broad_runner = broad.runner_from_config(config)
broad_runner.run_fn("picard_index_ref", ref_file)
for x in align_bams:
broad_runner.run_fn("picard_index", x)
coverage_depth = config["algorithm"].get("coverage_depth", "high").lower()
variant_regions = config["algorithm"].get("variant_regions", None)
confidence = "4.0" if coverage_depth in ["low"] else "30.0"
region = subset_variant_regions(variant_regions, region, out_file)
params = ["-R", ref_file,
"--standard_min_confidence_threshold_for_calling", confidence,
"--standard_min_confidence_threshold_for_emitting", confidence,
"--downsample_to_coverage", "250",
"--downsampling_type", "BY_SAMPLE",
]
for a in annotation.get_gatk_annotations(config):
params += ["--annotation", a]
for x in align_bams:
params += ["-I", x]
if dbsnp:
params += ["--dbsnp", dbsnp]
if region:
params += ["-L", bamprep.region_to_gatk(region), "--interval_set_rule", "INTERSECTION"]
return broad_runner, params
def _shared_gatk_call_prep(align_bams, items, ref_file, region, out_file, num_cores=1):
"""Shared preparation work for GATK variant calling.
"""
data = items[0]
config = data["config"]
broad_runner = broad.runner_from_config(config)
gatk_type = broad_runner.gatk_type()
for x in align_bams:
bam.index(x, config)
if _use_spark(num_cores, gatk_type):
# GATK4 spark runs use 2bit reference index
params = ["--reference", dd.get_ref_twobit(items[0])]
else:
picard_runner = broad.runner_from_path("picard", config)
picard_runner.run_fn("picard_index_ref", ref_file)
params = ["-R", ref_file]
coverage_depth_min = tz.get_in(["algorithm", "coverage_depth_min"], config)
if coverage_depth_min and coverage_depth_min < 4:
confidence = "4.0"
params += ["--standard_min_confidence_threshold_for_calling", confidence]
for a in annotation.get_gatk_annotations(config):
params += ["--annotation", a]
for x in align_bams:
params += ["-I", x]
variant_regions = bedutils.population_variant_regions(items)
region = subset_variant_regions(variant_regions, region, out_file, items)
if region:
if gatk_type == "gatk4":
params += ["-L", bamprep.region_to_gatk(region), "--interval-set-rule", "INTERSECTION"]
else:
params += ["-L", bamprep.region_to_gatk(region), "--interval_set_rule", "INTERSECTION"]
params += standard_cl_params(items)
return broad_runner, params
def _shared_gatk_call_prep(align_bams, items, ref_file, dbsnp, region, out_file):
"""Shared preparation work for GATK variant calling.
"""
config = items[0]["config"]
broad_runner = broad.runner_from_config(config)
broad_runner.run_fn("picard_index_ref", ref_file)
for x in align_bams:
bam.index(x, config)
# GATK can only downsample to a minimum of 200
coverage_depth_max = max(200, utils.get_in(config, ("algorithm", "coverage_depth_max"), 2000))
coverage_depth_min = utils.get_in(config, ("algorithm", "coverage_depth_min"), 4)
variant_regions = config["algorithm"].get("variant_regions", None)
confidence = "4.0" if coverage_depth_min < 4 else "30.0"
region = subset_variant_regions(variant_regions, region, out_file, items)
params = ["-R", ref_file,
"--standard_min_confidence_threshold_for_calling", confidence,
"--standard_min_confidence_threshold_for_emitting", confidence,
"--downsample_to_coverage", str(coverage_depth_max),
"--downsampling_type", "BY_SAMPLE",
]
for a in annotation.get_gatk_annotations(config):
params += ["--annotation", a]
for x in align_bams:
params += ["-I", x]
if dbsnp:
params += ["--dbsnp", dbsnp]
if region:
params += ["-L", bamprep.region_to_gatk(region), "--interval_set_rule", "INTERSECTION"]
return broad_runner, params
def _shared_gatk_call_prep(align_bams, ref_file, config, dbsnp, region, out_file):
"""Shared preparation work for GATK variant calling.
"""
broad_runner = broad.runner_from_config(config)
broad_runner.run_fn("picard_index_ref", ref_file)
for x in align_bams:
broad_runner.run_fn("picard_index", x)
coverage_depth = config["algorithm"].get("coverage_depth", "high").lower()
variant_regions = config["algorithm"].get("variant_regions", None)
confidence = "4.0" if coverage_depth in ["low"] else "30.0"
if out_file is None:
out_file = "%s-variants.vcf" % os.path.splitext(align_bams[0])[0]
region = subset_variant_regions(variant_regions, region, out_file)
params = ["-R", ref_file,
"--standard_min_confidence_threshold_for_calling", confidence,
"--standard_min_confidence_threshold_for_emitting", confidence,
]
for a in annotation.get_gatk_annotations(config):
params += ["--annotation", a]
for x in align_bams:
params += ["-I", x]
if dbsnp:
params += ["--dbsnp", dbsnp]
if region:
params += ["-L", region, "--interval_set_rule", "INTERSECTION"]
return broad_runner, params, out_file
def mutect2_caller(align_bams,
items,
ref_file,
assoc_files,
region=None,
out_file=None):
"""Call variation with GATK's MuTect2.
This requires the full non open-source version of GATK 3.5+.
"""
if out_file is None:
out_file = "%s-variants.vcf.gz" % utils.splitext_plus(align_bams[0])[0]
if not utils.file_exists(out_file):
paired = vcfutils.get_paired_bams(align_bams, items)
broad_runner = broad.runner_from_config(items[0]["config"])
gatk_type = broad_runner.gatk_type()
_prep_inputs(align_bams, ref_file, items)
with file_transaction(items[0], out_file) as tx_out_file:
params = [
"-T", "Mutect2" if gatk_type == "gatk4" else "MuTect2", "-R",
ref_file, "--annotation", "ClippingRankSumTest",
"--annotation", "DepthPerSampleHC"
]
for a in annotation.get_gatk_annotations(
items[0]["config"], include_baseqranksum=False):
params += ["--annotation", a]
# Avoid issues with BAM CIGAR reads that GATK doesn't like
if gatk_type == "gatk4":
params += ["--read-validation-stringency", "LENIENT"]
params += _add_tumor_params(paired, items, gatk_type)
params += _add_region_params(region, out_file, items, gatk_type)
# Avoid adding dbSNP/Cosmic so they do not get fed to variant filtering algorithm
# Not yet clear how this helps or hurts in a general case.
#params += _add_assoc_params(assoc_files)
params += ["-ploidy", str(ploidy.get_ploidy(items, region))]
resources = config_utils.get_resources("mutect2",
items[0]["config"])
if "options" in resources:
params += [str(x) for x in resources.get("options", [])]
assert LooseVersion(broad_runner.gatk_major_version()) >= LooseVersion("3.5"), \
"Require full version of GATK 3.5+ for mutect2 calling"
broad_runner.new_resources("mutect2")
gatk_cmd = broad_runner.cl_gatk(params,
os.path.dirname(tx_out_file))
if gatk_type == "gatk4":
tx_raw_prefilt_file = "%s-raw%s" % utils.splitext_plus(
tx_out_file)
tx_raw_file = "%s-raw-filt%s" % utils.splitext_plus(
tx_out_file)
filter_cmd = _mutect2_filter(broad_runner, tx_raw_prefilt_file,
tx_raw_file)
cmd = "{gatk_cmd} -O {tx_raw_prefilt_file} && {filter_cmd}"
else:
tx_raw_file = "%s-raw%s" % utils.splitext_plus(tx_out_file)
cmd = "{gatk_cmd} > {tx_raw_file}"
do.run(cmd.format(**locals()), "MuTect2")
out_file = _af_filter(paired.tumor_data, tx_raw_file, out_file)
return vcfutils.bgzip_and_index(out_file, items[0]["config"])
def mutect2_caller(align_bams, items, ref_file, assoc_files,
region=None, out_file=None):
"""Call variation with GATK's MuTect2.
This requires the full non open-source version of GATK 3.5+.
"""
if out_file is None:
out_file = "%s-variants.vcf.gz" % utils.splitext_plus(align_bams[0])[0]
if not utils.file_exists(out_file):
paired = vcfutils.get_paired_bams(align_bams, items)
broad_runner = broad.runner_from_config(items[0]["config"])
gatk_type = broad_runner.gatk_type()
_prep_inputs(align_bams, ref_file, items)
with file_transaction(items[0], out_file) as tx_out_file:
params = ["-T", "Mutect2" if gatk_type == "gatk4" else "MuTect2",
"--annotation", "ClippingRankSumTest",
"--annotation", "DepthPerSampleHC"]
if gatk_type == "gatk4":
params += ["--reference", ref_file]
else:
params += ["-R", ref_file]
for a in annotation.get_gatk_annotations(items[0]["config"], include_baseqranksum=False):
params += ["--annotation", a]
# Avoid issues with BAM CIGAR reads that GATK doesn't like
if gatk_type == "gatk4":
params += ["--read-validation-stringency", "LENIENT"]
params += _add_tumor_params(paired, items, gatk_type)
params += _add_region_params(region, out_file, items, gatk_type)
# Avoid adding dbSNP/Cosmic so they do not get fed to variant filtering algorithm
# Not yet clear how this helps or hurts in a general case.
#params += _add_assoc_params(assoc_files)
resources = config_utils.get_resources("mutect2", items[0]["config"])
if "options" in resources:
params += [str(x) for x in resources.get("options", [])]
assert LooseVersion(broad_runner.gatk_major_version()) >= LooseVersion("3.5"), \
"Require full version of GATK 3.5+ for mutect2 calling"
broad_runner.new_resources("mutect2")
gatk_cmd = broad_runner.cl_gatk(params, os.path.dirname(tx_out_file))
if gatk_type == "gatk4":
tx_raw_prefilt_file = "%s-raw%s" % utils.splitext_plus(tx_out_file)
tx_raw_file = "%s-raw-filt%s" % utils.splitext_plus(tx_out_file)
filter_cmd = _mutect2_filter(broad_runner, tx_raw_prefilt_file, tx_raw_file, ref_file)
cmd = "{gatk_cmd} -O {tx_raw_prefilt_file} && {filter_cmd}"
else:
tx_raw_file = "%s-raw%s" % utils.splitext_plus(tx_out_file)
cmd = "{gatk_cmd} > {tx_raw_file}"
do.run(cmd.format(**locals()), "MuTect2")
out_file = _af_filter(paired.tumor_data, tx_raw_file, out_file)
return vcfutils.bgzip_and_index(out_file, items[0]["config"])
def _shared_gatk_call_prep(align_bams, items, ref_file, dbsnp, cosmic, region, out_file):
"""Shared preparation work for GATK variant calling.
"""
data = items[0]
config = data["config"]
broad_runner = broad.runner_from_path("picard", config)
broad_runner.run_fn("picard_index_ref", ref_file)
for x in align_bams:
bam.index(x, config)
params = ["-R", ref_file]
coverage_depth_min = tz.get_in(["algorithm", "coverage_depth_min"], config)
if coverage_depth_min and coverage_depth_min < 4:
confidence = "4.0"
params += [
"--standard_min_confidence_threshold_for_calling",
confidence,
"--standard_min_confidence_threshold_for_emitting",
confidence,
]
for a in annotation.get_gatk_annotations(config):
params += ["--annotation", a]
for x in align_bams:
bam.index(x, config)
paired = vcfutils.get_paired_bams(align_bams, items)
if not paired:
raise ValueError(
"Specified MuTect calling but 'tumor' phenotype not present in batch\n"
"https://bcbio-nextgen.readthedocs.org/en/latest/contents/"
"pipelines.html#cancer-variant-calling\n"
"for samples: %s" % ", ".join([dd.get_sample_name(x) for x in items])
)
params += ["-I:tumor", paired.tumor_bam]
if paired.normal_bam is not None:
params += ["-I:normal", paired.normal_bam]
if paired.normal_panel is not None:
params += ["--normal_panel", paired.normal_panel]
if dbsnp:
params += ["--dbsnp", dbsnp]
if cosmic:
params += ["--cosmic", cosmic]
variant_regions = tz.get_in(["algorithm", "variant_regions"], config)
region = subset_variant_regions(variant_regions, region, out_file, items)
if region:
params += ["-L", bamprep.region_to_gatk(region), "--interval_set_rule", "INTERSECTION"]
broad_runner = broad.runner_from_config(config)
return broad_runner, params
def _shared_gatk_call_prep(align_bams, items, ref_file, dbsnp, cosmic, region, out_file):
"""Shared preparation work for GATK variant calling.
"""
data = items[0]
config = data["config"]
broad_runner = broad.runner_from_path("picard", config)
broad_runner.run_fn("picard_index_ref", ref_file)
for x in align_bams:
bam.index(x, config)
params = ["-R", ref_file]
coverage_depth_min = tz.get_in(["algorithm", "coverage_depth_min"], config)
if coverage_depth_min and coverage_depth_min < 4:
confidence = "4.0"
params += ["--standard_min_confidence_threshold_for_calling", confidence,
"--standard_min_confidence_threshold_for_emitting", confidence]
for a in annotation.get_gatk_annotations(config):
params += ["--annotation", a]
for x in align_bams:
bam.index(x, config)
paired = vcfutils.get_paired_bams(align_bams, items)
if not paired:
raise ValueError("Specified MuTect calling but 'tumor' phenotype not present in batch\n"
"https://bcbio-nextgen.readthedocs.org/en/latest/contents/"
"pipelines.html#cancer-variant-calling\n"
"for samples: %s" % ", " .join([dd.get_sample_name(x) for x in items]))
params += ["-I:tumor", paired.tumor_bam]
if paired.normal_bam is not None:
params += ["-I:normal", paired.normal_bam]
if paired.normal_panel is not None:
params += ["--normal_panel", paired.normal_panel]
if dbsnp:
params += ["--dbsnp", dbsnp]
if cosmic:
params += ["--cosmic", cosmic]
variant_regions = tz.get_in(["algorithm", "variant_regions"], config)
region = subset_variant_regions(variant_regions, region, out_file, items)
if region:
params += ["-L", bamprep.region_to_gatk(region), "--interval_set_rule", "INTERSECTION"]
broad_runner = broad.runner_from_config(config)
return broad_runner, params
def _shared_gatk_call_prep(align_bams, items, ref_file, dbsnp, region,
out_file):
"""Shared preparation work for GATK variant calling.
"""
data = items[0]
config = data["config"]
broad_runner = broad.runner_from_config(config)
broad_runner.run_fn("picard_index_ref", ref_file)
for x in align_bams:
bam.index(x, config)
params = ["-R", ref_file]
if dd.is_set_coverage_depth_max(data):
coverage_depth_max = dd.get_coverage_depth_max(data)
# GATK can only downsample to a minimum of 200
coverage_depth_max = max([200, coverage_depth_max])
params += [
"--downsample_to_coverage",
str(coverage_depth_max), "--downsampling_type", "BY_SAMPLE"
]
coverage_depth_min = tz.get_in(["algorithm", "coverage_depth_min"], config)
if coverage_depth_min and coverage_depth_min < 4:
confidence = "4.0"
params += [
"--standard_min_confidence_threshold_for_calling", confidence,
"--standard_min_confidence_threshold_for_emitting", confidence
]
for a in annotation.get_gatk_annotations(config):
params += ["--annotation", a]
for x in align_bams:
params += ["-I", x]
if dbsnp:
params += ["--dbsnp", dbsnp]
variant_regions = tz.get_in(["algorithm", "variant_regions"], config)
region = subset_variant_regions(variant_regions, region, out_file, items)
if region:
params += [
"-L",
bamprep.region_to_gatk(region), "--interval_set_rule",
"INTERSECTION"
]
return broad_runner, params
def mutect2_caller(align_bams, items, ref_file, assoc_files,
region=None, out_file=None):
"""Call variation with GATK's MuTect2.
This requires the full non open-source version of GATK 3.5+.
items = 1 sample or T/N pair
"""
if out_file is None:
out_file = "%s-variants.vcf.gz" % utils.splitext_plus(align_bams[0])[0]
if not utils.file_exists(out_file):
# call somatic variants keeping germline sites and using germline 1KG resource
# use --native-pair-hmm-threads?
broad_runner = broad.runner_from_config(items[0]["config"])
gatk_type = broad_runner.gatk_type()
# shared Mutect2 settings for PureCN analysis in the case of:
# - PON creation
# - Tumor-only PureCN run
# - T/N PureCN run
# PURECN requirement alters Mutect2 variants calling!
if "purecn" in dd.get_svcaller(items[0]):
# mutect call for PON creation or purecn T-only analysis
_prep_inputs(align_bams, ref_file, items)
with file_transaction(items[0], out_file) as tx_out_file:
germline_resource = tz.get_in(["genome_resources", "variation", "af_only_gnomad"], items[0])
germline_path = os.path.normpath(os.path.join(os.path.dirname(ref_file), germline_resource))
input_bam = dd.get_work_bam(items[0])
tx_prefilt_vcf = utils.splitext_plus(tx_out_file)[0] + ".prefilt.vcf"
tx_vcf = os.path.splitext(tx_out_file)[0]
out_file_ungz = os.path.splitext(out_file)[0]
params = ["-T", "Mutect2"]
# T/N pair
if len(items) == 2:
paired = vcfutils.get_paired_bams(align_bams, items)
# not really running purecn with mutect1/gatk3
params += _add_tumor_params(paired, items, gatk_type)
logger.debug("You are running mutect2 in PureCN analysis in T/N mode, T-only + PON is recommended")
else: #T only
params += ["-I", input_bam]
# adding SNV PON from background/variant
snv_pon = tz.get_in(["config", "algorithm", "background", "variant"], items[0])
if snv_pon and dd.get_batch(items[0]) != "pon_build":
params += ["-pon", snv_pon]
params += ["--genotype-pon-sites"]
opt_list = config_utils.get_resources("mutect2", items[0]["config"]).get("options")
# default is 50, sometimes 100 or 200 is recommended for better sensitivity in detection
# hom del CNVs (calling more variants helps)
interval_padding = 50
if opt_list:
opt_dict = dict(zip(opt_list[::2], opt_list[1::2]))
if "--interval_padding" in opt_dict:
interval_padding = opt_dict["--interval_padding"]
params += ["--max-mnp-distance", "0",
"--interval-padding", interval_padding,
"--germline-resource", germline_path,
"--genotype-germline-sites",
"--reference", ref_file,
"-O", tx_prefilt_vcf]
params += _add_region_params(region, out_file, items, gatk_type)
broad_runner.new_resources("mutect2")
gatk_cmd = broad_runner.cl_gatk(params, os.path.dirname(tx_out_file))
filter_cmd = _mutect2_filter(broad_runner, items, tx_prefilt_vcf, out_file_ungz, ref_file)
cmd = "{gatk_cmd} && {filter_cmd}"
do.run(cmd.format(**locals()), "MuTect2")
# no AF filter for PureCN variants
out_file = vcfutils.bgzip_and_index(out_file_ungz, items[0]["config"])
else:
# a regular mutect call
paired = vcfutils.get_paired_bams(align_bams, items)
f1r2_file = None
_prep_inputs(align_bams, ref_file, items)
with file_transaction(items[0], out_file) as tx_out_file:
params = ["-T", "Mutect2" if gatk_type == "gatk4" else "MuTect2",
"--annotation", "ClippingRankSumTest",
"--annotation", "DepthPerSampleHC"]
if gatk_type == "gatk4":
params += ["--reference", ref_file]
else:
params += ["-R", ref_file]
for a in annotation.get_gatk_annotations(items[0]["config"], include_baseqranksum=False):
params += ["--annotation", a]
# Avoid issues with BAM CIGAR reads that GATK doesn't like
if gatk_type == "gatk4":
params += ["--read-validation-stringency", "LENIENT"]
params += _add_tumor_params(paired, items, gatk_type)
params += _add_region_params(region, out_file, items, gatk_type)
if all(is_paired(bam) for bam in align_bams) and (
"mutect2_readmodel" in utils.get_in(items[0], "config", "tools_on")):
orientation_filter = True
else:
orientation_filter = False
if gatk_type == "gatk4" and orientation_filter:
f1r2_file = "{}-f1r2.tar.gz".format(utils.splitext_plus(out_file)[0])
params += ["--f1r2-tar-gz", f1r2_file]
# Avoid adding dbSNP/Cosmic so they do not get fed to variant filtering algorithm
# Not yet clear how this helps or hurts in a general case.
#params += _add_assoc_params(assoc_files)
resources = config_utils.get_resources("mutect2", items[0]["config"])
if "options" in resources:
params += [str(x) for x in resources.get("options", [])]
assert LooseVersion(broad_runner.gatk_major_version()) >= LooseVersion("3.5"), \
"Require full version of GATK 3.5+ for mutect2 calling"
broad_runner.new_resources("mutect2")
gatk_cmd = broad_runner.cl_gatk(params, os.path.dirname(tx_out_file))
if gatk_type == "gatk4":
tx_raw_prefilt_file = "%s-raw%s" % utils.splitext_plus(out_file)
tx_raw_file = "%s-raw-filt%s" % utils.splitext_plus(tx_out_file)
if orientation_filter:
tx_f1r2_file = "{}-read-orientation-model.tar.gz"
tx_f1r2_file = tx_f1r2_file.format(utils.splitext_plus(f1r2_file)[0])
tx_read_orient_cmd = _mutect2_read_filter(broad_runner,
f1r2_file,
tx_f1r2_file)
filter_cmd = _mutect2_filter(broad_runner, items,
tx_raw_prefilt_file,
tx_raw_file, ref_file,
tx_f1r2_file)
else:
filter_cmd = _mutect2_filter(broad_runner, items,
tx_raw_prefilt_file,
tx_raw_file, ref_file)
if orientation_filter:
cmd = "{gatk_cmd} -O {tx_raw_prefilt_file} && {tx_read_orient_cmd} && {filter_cmd}"
else:
cmd = "{gatk_cmd} -O {tx_raw_prefilt_file} && {filter_cmd}"
else:
tx_raw_file = "%s-raw%s" % utils.splitext_plus(tx_out_file)
cmd = "{gatk_cmd} > {tx_raw_file}"
do.run(cmd.format(**locals()), "MuTect2")
out_file = _af_filter(paired.tumor_data, tx_raw_file, out_file)
return vcfutils.bgzip_and_index(out_file, items[0]["config"])
