def run(items):
paired = vcfutils.get_paired(items)
# paired is PairedInfo of one T/N pair (or just T) - named tuple, paired.tumor_config
if not paired:
logger.info("Skipping PureCN; no somatic tumor calls in batch: %s" %
" ".join([dd.get_sample_name(d) for d in items]))
return items
work_dir = _sv_workdir(paired.tumor_data)
normaldb = tz.get_in(["algorithm", "background", "cnv_reference", "purecn_normaldb"], paired.tumor_config)
# the right way of running purecn is with normaldb
if normaldb:
purecn_out = _run_purecn_normaldb(paired, work_dir)
purecn_out = _run_purecn_dx(purecn_out, paired)
else:
purecn_out = _run_purecn(paired, work_dir)
out = []
if paired.normal_data:
out.append(paired.normal_data)
if purecn_out:
purecn_out["variantcaller"] = "purecn"
if "loh" in purecn_out:
from bcbio.structural import titancna
purecn_out["vrn_file"] = titancna.to_vcf(purecn_out["loh"], "PureCN", _get_header, _loh_to_vcf,
paired.tumor_data, sep=",")
purecn_out["lohsummary"] = loh.summary_status(purecn_out, paired.tumor_data)
if "sv" not in paired.tumor_data:
paired.tumor_data["sv"] = []
paired.tumor_data["sv"].append(purecn_out)
out.append(paired.tumor_data)
return out
def _run_paired(paired):
"""Run somatic variant calling pipeline.
"""
from bcbio.structural import titancna
work_dir = _sv_workdir(paired.tumor_data)
seg_files = model_segments(
tz.get_in(["depth", "bins", "normalized"], paired.tumor_data),
work_dir, paired)
call_file = call_copy_numbers(seg_files["seg"], work_dir,
paired.tumor_data)
out = []
if paired.normal_data:
out.append(paired.normal_data)
if "sv" not in paired.tumor_data:
paired.tumor_data["sv"] = []
paired.tumor_data["sv"].append({
"variantcaller":
"gatk-cnv",
"call_file":
call_file,
"vrn_file":
titancna.to_vcf(call_file, "GATK4-CNV", _get_seg_header, _seg_to_vcf,
paired.tumor_data),
"seg":
seg_files["seg"],
"plot":
plot_model_segments(seg_files, work_dir, paired.tumor_data)
})
out.append(paired.tumor_data)
return out
def run(items):
paired = vcfutils.get_paired(items)
if not paired:
logger.info("Skipping PureCN; no somatic tumor calls in batch: %s" %
" ".join([dd.get_sample_name(d) for d in items]))
return items
work_dir = _sv_workdir(paired.tumor_data)
purecn_out = _run_purecn(paired, work_dir)
# XXX Currently finding edge case failures with Dx calling, needs additional testing
# purecn_out = _run_purecn_dx(purecn_out, paired)
out = []
if paired.normal_data:
out.append(paired.normal_data)
if purecn_out:
purecn_out["variantcaller"] = "purecn"
if "loh" in purecn_out:
from bcbio.structural import titancna
purecn_out["vrn_file"] = titancna.to_vcf(purecn_out["loh"], "PureCN", _get_header, _loh_to_vcf,
paired.tumor_data, sep=",")
purecn_out["lohsummary"] = loh.summary_status(purecn_out, paired.tumor_data)
if "sv" not in paired.tumor_data:
paired.tumor_data["sv"] = []
paired.tumor_data["sv"].append(purecn_out)
out.append(paired.tumor_data)
return out
def run(items):
paired = vcfutils.get_paired(items)
if not paired:
logger.info("Skipping PureCN; no somatic tumor calls in batch: %s" %
" ".join([dd.get_sample_name(d) for d in items]))
return items
work_dir = _sv_workdir(paired.tumor_data)
purecn_out = _run_purecn(paired, work_dir)
# XXX Currently finding edge case failures with Dx calling, needs additional testing
# purecn_out = _run_purecn_dx(purecn_out, paired)
out = []
if paired.normal_data:
out.append(paired.normal_data)
if purecn_out:
purecn_out["variantcaller"] = "purecn"
if "loh" in purecn_out:
from bcbio.structural import titancna
purecn_out["vrn_file"] = titancna.to_vcf(purecn_out["loh"],
"PureCN",
_get_header,
_loh_to_vcf,
paired.tumor_data,
sep=",")
purecn_out["lohsummary"] = loh.summary_status(
purecn_out, paired.tumor_data)
if "sv" not in paired.tumor_data:
paired.tumor_data["sv"] = []
paired.tumor_data["sv"].append(purecn_out)
out.append(paired.tumor_data)
return out
def _run_purple(paired, het_file, depth_file, vrn_files, work_dir):
"""Run PURPLE with pre-calculated AMBER and COBALT compatible inputs.
"""
purple_dir = utils.safe_makedir(os.path.join(work_dir, "purple"))
out_file = os.path.join(purple_dir, "%s.purple.cnv" % dd.get_sample_name(paired.tumor_data))
if not utils.file_exists(out_file):
with file_transaction(paired.tumor_data, out_file) as tx_out_file:
cmd = ["PURPLE"] + _get_jvm_opts(tx_out_file, paired.tumor_data) + \
["-amber", os.path.dirname(het_file), "-baf", het_file,
"-cobalt", os.path.dirname(depth_file),
"-gc_profile", dd.get_variation_resources(paired.tumor_data)["gc_profile"],
"-output_dir", os.path.dirname(tx_out_file),
"-ref_genome", "hg38" if dd.get_genome_build(paired.tumor_data) == "hg38" else "hg19",
"-run_dir", work_dir,
"-threads", dd.get_num_cores(paired.tumor_data),
"-tumor_sample", dd.get_sample_name(paired.tumor_data),
"-ref_sample", dd.get_sample_name(paired.normal_data)]
if vrn_files:
cmd += ["-somatic_vcf", vrn_files[0]["vrn_file"]]
# Avoid X11 display errors when writing plots
cmd = "unset DISPLAY && %s" % " ".join([str(x) for x in cmd])
do.run(cmd, "PURPLE: purity and ploidy estimation")
for f in os.listdir(os.path.dirname(tx_out_file)):
if f != os.path.basename(tx_out_file):
shutil.move(os.path.join(os.path.dirname(tx_out_file), f),
os.path.join(purple_dir, f))
out_file_export = os.path.join(purple_dir, "%s-purple-cnv.tsv" % (dd.get_sample_name(paired.tumor_data)))
if not utils.file_exists(out_file_export):
utils.symlink_plus(out_file, out_file_export)
out = {"variantcaller": "purple", "call_file": out_file_export,
"vrn_file": titancna.to_vcf(out_file_export, "PURPLE", _get_header, _export_to_vcf,
paired.tumor_data),
"plot": {}, "metrics": {}}
for name, ext in [("copy_number", "copyNumber"), ("minor_allele", "minor_allele"), ("variant", "variant")]:
plot_file = os.path.join(purple_dir, "plot", "%s.%s.png" % (dd.get_sample_name(paired.tumor_data), ext))
if os.path.exists(plot_file):
out["plot"][name] = plot_file
purity_file = os.path.join(purple_dir, "%s.purple.purity" % dd.get_sample_name(paired.tumor_data))
with open(purity_file) as in_handle:
header = in_handle.readline().replace("#", "").split("\t")
vals = in_handle.readline().split("\t")
for h, v in zip(header, vals):
try:
v = float(v)
except ValueError:
pass
out["metrics"][h] = v
return out
def _run_paired(paired):
"""Run somatic variant calling pipeline.
"""
from bcbio.structural import titancna
work_dir = _sv_workdir(paired.tumor_data)
seg_files = model_segments(tz.get_in(["depth", "bins", "normalized"], paired.tumor_data),
work_dir, paired)
call_file = call_copy_numbers(seg_files["seg"], work_dir, paired.tumor_data)
out = []
if paired.normal_data:
out.append(paired.normal_data)
if "sv" not in paired.tumor_data:
paired.tumor_data["sv"] = []
paired.tumor_data["sv"].append({"variantcaller": "gatk-cnv",
"call_file": call_file,
"vrn_file": titancna.to_vcf(call_file, "GATK4-CNV", _get_seg_header,
_seg_to_vcf, paired.tumor_data),
"seg": seg_files["seg"],
"plot": plot_model_segments(seg_files, work_dir, paired.tumor_data)})
out.append(paired.tumor_data)
return out
def _run_purple(paired, het_file, depth_file, vrn_files, work_dir):
"""Run PURPLE with pre-calculated AMBER and COBALT compatible inputs.
"""
purple_dir = utils.safe_makedir(os.path.join(work_dir, "purple"))
out_file = os.path.join(
purple_dir, "%s.purple.cnv" % dd.get_sample_name(paired.tumor_data))
if not utils.file_exists(out_file):
with file_transaction(paired.tumor_data, out_file) as tx_out_file:
cmd = ["PURPLE"] + _get_jvm_opts(tx_out_file, paired.tumor_data) + \
["-amber", os.path.dirname(het_file), "-baf", het_file,
"-cobalt", os.path.dirname(depth_file),
"-gc_profile", dd.get_variation_resources(paired.tumor_data)["gc_profile"],
"-output_dir", os.path.dirname(tx_out_file),
"-ref_genome", "hg38" if dd.get_genome_build(paired.tumor_data) == "hg38" else "hg19",
"-run_dir", work_dir,
"-threads", dd.get_num_cores(paired.tumor_data),
"-tumor_sample", dd.get_sample_name(paired.tumor_data),
"-ref_sample", dd.get_sample_name(paired.normal_data)]
if vrn_files:
cmd += ["-somatic_vcf", vrn_files[0]["vrn_file"]]
# Avoid X11 display errors when writing plots
cmd = "unset DISPLAY && %s" % " ".join([str(x) for x in cmd])
do.run(cmd, "PURPLE: purity and ploidy estimation")
for f in os.listdir(os.path.dirname(tx_out_file)):
if f != os.path.basename(tx_out_file):
shutil.move(os.path.join(os.path.dirname(tx_out_file), f),
os.path.join(purple_dir, f))
out_file_export = os.path.join(
purple_dir,
"%s-purple-cnv.tsv" % (dd.get_sample_name(paired.tumor_data)))
if not utils.file_exists(out_file_export):
utils.symlink_plus(out_file, out_file_export)
out = {
"variantcaller":
"purple",
"call_file":
out_file_export,
"vrn_file":
titancna.to_vcf(out_file_export, "PURPLE", _get_header, _export_to_vcf,
paired.tumor_data),
"plot": {},
"metrics": {}
}
for name, ext in [("copy_number", "copyNumber"),
("minor_allele", "minor_allele"),
("variant", "variant")]:
plot_file = os.path.join(
purple_dir, "plot",
"%s.%s.png" % (dd.get_sample_name(paired.tumor_data), ext))
if os.path.exists(plot_file):
out["plot"][name] = plot_file
purity_file = os.path.join(
purple_dir, "%s.purple.purity" % dd.get_sample_name(paired.tumor_data))
with open(purity_file) as in_handle:
header = in_handle.readline().replace("#", "").split("\t")
vals = in_handle.readline().split("\t")
for h, v in zip(header, vals):
try:
v = float(v)
except ValueError:
pass
out["metrics"][h] = v
return out
