def _produce_compatible_vcf(out_file, data, is_somatic):
"""Create a compatible VCF that downstream tools can deal with.
- htsjdk and thus GATK and Picard do not support VCF4.3:
https://github.com/broadinstitute/gatk/issues/2092
- Use octopus legacy format to avoid incompatibilities.
https://github.com/luntergroup/octopus#output-format
- Fixes `##contig` lines since octopus only writes contigs
used in the BED file region, causing incompatibilies with
GatherVcfs when merging
- Fixes alleles prefixed with '*' like 'C,*T' which cause
downstream failures when reading with GATK.
"""
base, ext = utils.splitext_plus(out_file)
legacy_file = "%s.legacy%s" % (base, ext)
if is_somatic:
legacy_file = _covert_to_diploid(legacy_file, data)
final_file = "%s.vcf.gz" % base
cat_cmd = "zcat" if legacy_file.endswith(".gz") else "cat"
contig_cl = vcfutils.add_contig_to_header_cl(dd.get_ref_file(data), out_file)
remove_problem_alleles = r"sed 's/,\*\([A-Z]\)/,\1/'"
cmd = ("{cat_cmd} {legacy_file} | sed 's/fileformat=VCFv4.3/fileformat=VCFv4.2/' | "
"{remove_problem_alleles} | {contig_cl} | bgzip -c > {final_file}")
do.run(cmd.format(**locals()), "Produce compatible VCF output file from octopus")
return vcfutils.bgzip_and_index(out_file, data["config"])
def run(align_bams, items, ref_file, assoc_files, region=None, out_file=None):
"""Run tumor only smCounter2 calling.
"""
paired = vcfutils.get_paired_bams(align_bams, items)
assert paired and not paired.normal_bam, ("Pisces supports tumor-only variant calling: %s" %
(",".join([dd.get_sample_name(d) for d in items])))
vrs = bedutils.population_variant_regions(items)
target = shared.subset_variant_regions(vrs, region,
out_file, items=items, do_merge=True)
out_file = out_file.replace(".vcf.gz", ".vcf")
out_prefix = utils.splitext_plus(os.path.basename(out_file))[0]
if not utils.file_exists(out_file) and not utils.file_exists(out_file + ".gz"):
with file_transaction(paired.tumor_data, out_file) as tx_out_file:
cmd = ["smCounter2", "--runPath", os.path.dirname(tx_out_file),
"--outPrefix", out_prefix,
"--bedTarget", target, "--refGenome", ref_file,
"--bamFile", paired.tumor_bam, "--bamType", "consensus",
"--nCPU", dd.get_num_cores(paired.tumor_data)]
do.run(cmd, "smcounter2 variant calling")
for fname in glob.glob(os.path.join(os.path.dirname(tx_out_file), "*.smCounter*")):
shutil.move(fname, os.path.join(os.path.dirname(out_file), os.path.basename(fname)))
utils.symlink_plus(os.path.join(os.path.dirname(out_file),
"%s.smCounter.cut.vcf" % out_prefix),
out_file)
return vcfutils.bgzip_and_index(out_file, paired.tumor_data["config"], remove_orig=False,
prep_cmd="sed 's#FORMAT\t%s#FORMAT\t%s#' | %s" %
(out_prefix, dd.get_sample_name(paired.tumor_data),
vcfutils.add_contig_to_header_cl(dd.get_ref_file(paired.tumor_data), out_file)))
def _produce_compatible_vcf(out_file, data, is_somatic=False):
"""Create a compatible VCF that downstream tools can deal with.
- htsjdk and thus GATK and Picard do not support VCF4.3:
https://github.com/broadinstitute/gatk/issues/2092
- Use octopus legacy format to avoid incompatibilities.
https://github.com/luntergroup/octopus#output-format
- Fixes `##contig` lines since octopus only writes contigs
used in the BED file region, causing incompatibilies with
GatherVcfs when merging
- Fixes alleles prefixed with '*' like 'C,*T' which cause
downstream failures when reading with GATK.
- Changes phase set (PS) header to be type Integer.
"""
base, ext = utils.splitext_plus(out_file)
#legacy_file = "%s.legacy%s" % (base, ext)
legacy_file = out_file
if is_somatic:
legacy_file = _covert_to_diploid(legacy_file, data)
final_file = "%s.vcf.gz" % base
cat_cmd = "zcat" if legacy_file.endswith(".gz") else "cat"
contig_cl = vcfutils.add_contig_to_header_cl(dd.get_ref_file(data),
out_file)
remove_problem_alleles = r"sed 's/,\*\([A-Z]\)/,\1/'"
fix_phasing_header = r"sed 's/ID=PS,Number=1,Type=String/ID=PS,Number=1,Type=Integer/'"
cmd = (
"{cat_cmd} {legacy_file} | sed 's/fileformat=VCFv4.3/fileformat=VCFv4.2/' | "
"{remove_problem_alleles} | {fix_phasing_header} | {contig_cl} | bgzip -c > {final_file}"
)
do.run(cmd.format(**locals()),
"Produce compatible VCF output file from octopus")
return vcfutils.bgzip_and_index(out_file, data["config"])
def _run_scalpel_paired(align_bams, items, ref_file, assoc_files,
region=None, out_file=None):
"""Detect indels with Scalpel.
This is used for paired tumor / normal samples.
"""
config = items[0]["config"]
if out_file is None:
out_file = "%s-paired-variants.vcf.gz" % os.path.splitext(align_bams[0])[0]
if not utils.file_exists(out_file):
with file_transaction(config, out_file) as tx_out_file:
paired = get_paired_bams(align_bams, items)
if not paired.normal_bam:
ann_file = _run_scalpel_caller(align_bams, items, ref_file,
assoc_files, region, out_file)
return ann_file
vcfstreamsort = config_utils.get_program("vcfstreamsort", config)
perl_exports = utils.get_perl_exports(os.path.dirname(tx_out_file))
tmp_path = "%s-scalpel-work" % utils.splitext_plus(out_file)[0]
db_file = os.path.join(tmp_path, "main", "somatic.db")
if not os.path.exists(db_file + ".dir"):
if os.path.exists(tmp_path):
utils.remove_safe(tmp_path)
opts = " ".join(_scalpel_options_from_config(items, config, out_file, region, tmp_path))
opts += " --ref {}".format(ref_file)
opts += " --dir %s" % tmp_path
# caling
cl = ("{perl_exports} && "
"scalpel-discovery --somatic {opts} --tumor {paired.tumor_bam} --normal {paired.normal_bam}")
do.run(cl.format(**locals()), "Genotyping paired variants with Scalpel", {})
# filtering to adjust input parameters
bed_opts = " ".join(_scalpel_bed_file_opts(items, config, out_file, region, tmp_path))
use_defaults = True
if use_defaults:
scalpel_tmp_file = os.path.join(tmp_path, "main/somatic.indel.vcf")
# Uses default filters but can tweak min-alt-count-tumor and min-phred-fisher
# to swap precision for sensitivity
else:
scalpel_tmp_file = os.path.join(tmp_path, "main/somatic-indel-filter.vcf.gz")
with file_transaction(config, scalpel_tmp_file) as tx_indel_file:
cmd = ("{perl_exports} && "
"scalpel-export --somatic {bed_opts} --ref {ref_file} --db {db_file} "
"--min-alt-count-tumor 5 --min-phred-fisher 10 --min-vaf-tumor 0.1 "
"| bgzip -c > {tx_indel_file}")
do.run(cmd.format(**locals()), "Scalpel somatic indel filter", {})
scalpel_tmp_file = bgzip_and_index(scalpel_tmp_file, config)
scalpel_tmp_file_common = bgzip_and_index(os.path.join(tmp_path, "main/common.indel.vcf"), config)
compress_cmd = "| bgzip -c" if out_file.endswith("gz") else ""
bcftools_cmd_chi2 = get_scalpel_bcftools_filter_expression("chi2", config)
bcftools_cmd_common = get_scalpel_bcftools_filter_expression("reject", config)
fix_ambig = vcfutils.fix_ambiguous_cl()
add_contig = vcfutils.add_contig_to_header_cl(dd.get_ref_file(items[0]), tx_out_file)
cl2 = ("vcfcat <({bcftools_cmd_chi2} {scalpel_tmp_file}) "
"<({bcftools_cmd_common} {scalpel_tmp_file_common}) | "
" {fix_ambig} | {vcfstreamsort} | {add_contig} {compress_cmd} > {tx_out_file}")
do.run(cl2.format(**locals()), "Finalising Scalpel variants", {})
return out_file
def _add_contig_cl(in_file, items):
has_contigs = False
with utils.open_gzipsafe(in_file) as in_handle:
for line in in_handle:
if line.startswith("##contig"):
has_contigs = True
break
elif not line.startswith("##"):
break
if not has_contigs:
return vcfutils.add_contig_to_header_cl(items[0])
def _run_scalpel_caller(align_bams,
items,
ref_file,
assoc_files,
region=None,
out_file=None):
"""Detect indels with Scalpel.
Single sample mode.
"""
config = items[0]["config"]
if out_file is None:
out_file = "%s-variants.vcf.gz" % os.path.splitext(align_bams[0])[0]
if not utils.file_exists(out_file):
with file_transaction(config, out_file) as tx_out_file:
if len(align_bams) > 1:
message = ("Scalpel does not currently support batch calling!")
raise ValueError(message)
input_bams = " ".join("%s" % x for x in align_bams)
tmp_path = "%s-scalpel-work" % utils.splitext_plus(out_file)[0]
tx_tmp_path = "%s-scalpel-work" % utils.splitext_plus(
tx_out_file)[0]
if os.path.exists(tmp_path):
utils.remove_safe(tmp_path)
opts = " ".join(
_scalpel_options_from_config(items, config, out_file, region,
tmp_path))
opts += " --dir %s" % tx_tmp_path
min_cov = "3" # minimum coverage
opts += " --mincov %s" % min_cov
perl_exports = utils.get_perl_exports(os.path.dirname(tx_out_file))
cmd = (
"{perl_exports} && "
"scalpel-discovery --single {opts} --ref {ref_file} --bam {input_bams} "
)
do.run(cmd.format(**locals()), "Genotyping with Scalpel", {})
shutil.move(tx_tmp_path, tmp_path)
# parse produced variant file further
scalpel_tmp_file = bgzip_and_index(
os.path.join(tmp_path, "variants.indel.vcf"), config)
compress_cmd = "| bgzip -c" if out_file.endswith("gz") else ""
bcftools_cmd_chi2 = get_scalpel_bcftools_filter_expression(
"chi2", config)
sample_name_str = items[0]["name"][1]
fix_ambig = vcfutils.fix_ambiguous_cl()
add_contig = vcfutils.add_contig_to_header_cl(
dd.get_ref_file(items[0]), tx_out_file)
cl2 = (
"{bcftools_cmd_chi2} {scalpel_tmp_file} | "
r"sed 's/FORMAT\tsample\(_name\)\{{0,1\}}/FORMAT\t{sample_name_str}/g' "
"| {fix_ambig} | vcfallelicprimitives -t DECOMPOSED --keep-geno | vcffixup - | vcfstreamsort "
"| {add_contig} {compress_cmd} > {tx_out_file}")
do.run(cl2.format(**locals()), "Finalising Scalpel variants", {})
return out_file
def _add_contig_cl(in_file, items):
has_contigs = False
with utils.open_gzipsafe(in_file) as in_handle:
for line in in_handle:
if line.startswith("##contig"):
has_contigs = True
break
elif not line.startswith("##"):
break
if not has_contigs:
return vcfutils.add_contig_to_header_cl(items[0])
def map_coords_to_ucsc(grc_cosmic, ref_file, out_file):
hg19_ref_file = ref_file.replace("GRCh37", "hg19")
if not os.path.exists(out_file):
contig_cl = vcfutils.add_contig_to_header_cl(hg19_ref_file, out_file)
cmd = ("zcat {grc_cosmic} | "
r'sed "s/^\([0-9]\+\)\t/chr\1\t/g" | sed "s/^MT/chrM/g" | sed "s/^X/chrX/g" | sed "s/^Y/chrY/g" '
"| {contig_cl} "
"| bgzip -c > {out_file}")
subprocess.check_call(cmd.format(**locals()), shell=True)
if os.path.exists("%s-header.txt" % utils.splitext_plus(out_file)[0]):
os.remove("%s-header.txt" % utils.splitext_plus(out_file)[0])
return vcfutils.bgzip_and_index(out_file, {})
def map_coords_to_ucsc(grc_cosmic, ref_file, out_file):
hg19_ref_file = ref_file.replace("GRCh37", "hg19")
if not os.path.exists(out_file):
contig_cl = vcfutils.add_contig_to_header_cl(hg19_ref_file, out_file)
cmd = ("zcat {grc_cosmic} | "
r'sed "s/^\([0-9]\+\)\t/chr\1\t/g" | sed "s/^MT/chrM/g" | sed "s/^X/chrX/g" | sed "s/^Y/chrY/g" '
"| {contig_cl} "
"| bgzip -c > {out_file}")
subprocess.check_call(cmd.format(**locals()), shell=True)
if os.path.exists("%s-header.txt" % utils.splitext_plus(out_file)[0]):
os.remove("%s-header.txt" % utils.splitext_plus(out_file)[0])
return vcfutils.bgzip_and_index(out_file, {})
def run(align_bams, items, ref_file, assoc_files, region, out_file):
"""Run platypus variant calling, germline whole genome or exome.
"""
assert out_file.endswith(".vcf.gz")
if not utils.file_exists(out_file):
with file_transaction(items[0], out_file) as tx_out_file:
for align_bam in align_bams:
bam.index(align_bam, items[0]["config"])
cmd = [
"platypus", "callVariants",
"--regions=%s" % _subset_regions(region, out_file, items),
"--bamFiles=%s" % ",".join(align_bams),
"--refFile=%s" % dd.get_ref_file(items[0]), "--output=-",
"--logFileName", "/dev/null", "--verbosity=1"
]
resources = config_utils.get_resources("platypus",
items[0]["config"])
if resources.get("options"):
# normalize options so we can set defaults without overwriting user specified
for opt in resources["options"]:
if "=" in opt:
key, val = opt.split("=")
cmd.extend([key, val])
else:
cmd.append(opt)
if any("gvcf" in dd.get_tools_on(d) for d in items):
cmd += ["--outputRefCalls", "1", "--refCallBlockSize", "50000"]
# Adjust default filter thresholds to achieve similar sensitivity/specificity to other callers
# Currently not used after doing more cross validation as they increase false positives
# which seems to be a major advantage for Platypus users.
# tuned_opts = ["--hapScoreThreshold", "10", "--scThreshold", "0.99", "--filteredReadsFrac", "0.9",
# "--rmsmqThreshold", "20", "--qdThreshold", "0", "--abThreshold", "0.0001",
# "--minVarFreq", "0.0", "--assemble", "1"]
# for okey, oval in utils.partition_all(2, tuned_opts):
# if okey not in cmd:
# cmd.extend([okey, oval])
# Avoid filtering duplicates on high depth targeted regions where we don't mark duplicates
if any(not dd.get_mark_duplicates(data) for data in items):
cmd += ["--filterDuplicates=0"]
post_process_cmd = (
" | %s | %s | %s | vcfallelicprimitives -t DECOMPOSED --keep-geno | vcffixup - | "
"vcfstreamsort | bgzip -c > %s" %
(vcfutils.fix_ambiguous_cl(), vcfutils.fix_ambiguous_cl(5),
vcfutils.add_contig_to_header_cl(dd.get_ref_file(items[0]),
tx_out_file), tx_out_file))
do.run(" ".join(cmd) + post_process_cmd,
"platypus variant calling")
out_file = vcfutils.bgzip_and_index(out_file, items[0]["config"])
return out_file
def sort_to_ref(fname, ref_file, add_chr):
"""Match reference genome ordering.
"""
out_file = "%s-prep.vcf.gz" % (fname.replace(".vcf.gz", ""))
if not os.path.exists(out_file):
if add_chr:
fix_chrom = r'| sed "s/^\([0-9]\+\)\t/chr\1\t/g" | sed "s/^MT/chrM/g" | sed "s/^X/chrX/g" | sed "s/^Y/chrY/g" '
else:
fix_chrom = ''
contig_cl = vcfutils.add_contig_to_header_cl(ref_file, out_file)
cmd = ("gunzip -c {fname} {fix_chrom} | "
"gsort /dev/stdin {ref_file}.fai | {contig_cl} | "
"bgzip -c > {out_file}")
subprocess.check_call(cmd.format(**locals()), shell=True)
return vcfutils.bgzip_and_index(out_file, {})
def sort_to_ref(fname, ref_file, add_chr):
"""Match reference genome ordering.
"""
out_file = "%s-prep.vcf.gz" % (fname.replace(".vcf.gz", ""))
if not os.path.exists(out_file):
if add_chr:
fix_chrom = r'| sed "s/^\([0-9]\+\)\t/chr\1\t/g" | sed "s/^MT/chrM/g" | sed "s/^X/chrX/g" | sed "s/^Y/chrY/g" '
else:
fix_chrom = ''
contig_cl = vcfutils.add_contig_to_header_cl(ref_file, out_file)
cmd = ("gunzip -c {fname} {fix_chrom} | "
"bcftools norm --check-ref s --do-not-normalize -f {ref_file} |"
"gsort /dev/stdin {ref_file}.fai | {contig_cl} | "
"bgzip -c > {out_file}")
subprocess.check_call(cmd.format(**locals()), shell=True)
return vcfutils.bgzip_and_index(out_file, {})
def run(align_bams, items, ref_file, assoc_files, region, out_file):
"""Run platypus variant calling, germline whole genome or exome.
"""
assert out_file.endswith(".vcf.gz")
if not utils.file_exists(out_file):
with file_transaction(items[0], out_file) as tx_out_file:
for align_bam in align_bams:
bam.index(align_bam, items[0]["config"])
cmd = ["platypus", "callVariants", "--regions=%s" % _subset_regions(region, out_file, items),
"--bamFiles=%s" % ",".join(align_bams),
"--refFile=%s" % dd.get_ref_file(items[0]), "--output=-",
"--logFileName", "/dev/null", "--verbosity=1"]
resources = config_utils.get_resources("platypus", items[0]["config"])
if resources.get("options"):
# normalize options so we can set defaults without overwriting user specified
for opt in resources["options"]:
if "=" in opt:
key, val = opt.split("=")
cmd.extend([key, val])
else:
cmd.append(opt)
if any("gvcf" in dd.get_tools_on(d) for d in items):
cmd += ["--outputRefCalls", "1", "--refCallBlockSize", "50000"]
# Adjust default filter thresholds to achieve similar sensitivity/specificity to other callers
# Currently not used after doing more cross validation as they increase false positives
# which seems to be a major advantage for Platypus users.
# tuned_opts = ["--hapScoreThreshold", "10", "--scThreshold", "0.99", "--filteredReadsFrac", "0.9",
# "--rmsmqThreshold", "20", "--qdThreshold", "0", "--abThreshold", "0.0001",
# "--minVarFreq", "0.0", "--assemble", "1"]
# for okey, oval in utils.partition_all(2, tuned_opts):
# if okey not in cmd:
# cmd.extend([okey, oval])
# Avoid filtering duplicates on high depth targeted regions where we don't mark duplicates
if any(not dd.get_mark_duplicates(data) for data in items):
cmd += ["--filterDuplicates=0"]
post_process_cmd = (" | %s | %s | %s | vcfallelicprimitives -t DECOMPOSED --keep-geno | vcffixup - | "
"vcfstreamsort | bgzip -c > %s" % (vcfutils.fix_ambiguous_cl(),
vcfutils.fix_ambiguous_cl(5),
vcfutils.add_contig_to_header_cl(items[0]),
tx_out_file))
do.run(" ".join(cmd) + post_process_cmd, "platypus variant calling")
out_file = vcfutils.bgzip_and_index(out_file, items[0]["config"])
return out_file
def sort_to_ref(fname, ref_file, add_chr):
"""Match reference genome ordering.
"""
logging.info(f"Sorting {fname} to match the order of {ref_file}.")
out_file = "%s-prep.vcf.gz" % (fname.replace(".vcf.gz", ""))
if not os.path.exists(out_file):
if add_chr:
fix_chrom = r'| sed "s/^\([0-9]\+\)\t/chr\1\t/g" | sed "s/^MT/chrM/g" | sed "s/^X/chrX/g" | sed "s/^Y/chrY/g" '
else:
fix_chrom = ''
contig_cl = vcfutils.add_contig_to_header_cl(ref_file, out_file)
cmd = ("gunzip -c {fname} {fix_chrom} | "
"bcftools norm --check-ref s --do-not-normalize -f {ref_file} |"
"bcftools view -e 'SNP=1' |"
"gsort /dev/stdin {ref_file}.fai | {contig_cl} | "
"bgzip -c > {out_file}")
subprocess.check_call(cmd.format(**locals()), shell=True)
logging.info(f"bgzipping and indexing {out_file}.")
return vcfutils.bgzip_and_index(out_file, {})
def _run_scalpel_caller(align_bams, items, ref_file, assoc_files,
region=None, out_file=None):
"""Detect indels with Scalpel.
Single sample mode.
"""
config = items[0]["config"]
if out_file is None:
out_file = "%s-variants.vcf.gz" % os.path.splitext(align_bams[0])[0]
if not utils.file_exists(out_file):
with file_transaction(config, out_file) as tx_out_file:
if len(align_bams) > 1:
message = ("Scalpel does not currently support batch calling!")
raise ValueError(message)
input_bams = " ".join("%s" % x for x in align_bams)
tmp_path = "%s-scalpel-work" % utils.splitext_plus(out_file)[0]
tx_tmp_path = "%s-scalpel-work" % utils.splitext_plus(tx_out_file)[0]
if os.path.exists(tmp_path):
utils.remove_safe(tmp_path)
opts = " ".join(_scalpel_options_from_config(items, config, out_file, region, tmp_path))
opts += " --dir %s" % tx_tmp_path
min_cov = "3" # minimum coverage
opts += " --mincov %s" % min_cov
perl_exports = utils.get_perl_exports(os.path.dirname(tx_out_file))
cmd = ("{perl_exports} && "
"scalpel-discovery --single {opts} --ref {ref_file} --bam {input_bams} ")
do.run(cmd.format(**locals()), "Genotyping with Scalpel", {})
shutil.move(tx_tmp_path, tmp_path)
# parse produced variant file further
scalpel_tmp_file = bgzip_and_index(os.path.join(tmp_path, "variants.indel.vcf"), config)
compress_cmd = "| bgzip -c" if out_file.endswith("gz") else ""
bcftools_cmd_chi2 = get_scalpel_bcftools_filter_expression("chi2", config)
sample_name_str = items[0]["name"][1]
fix_ambig = vcfutils.fix_ambiguous_cl()
add_contig = vcfutils.add_contig_to_header_cl(dd.get_ref_file(items[0]), tx_out_file)
cl2 = ("{bcftools_cmd_chi2} {scalpel_tmp_file} | "
r"sed 's/FORMAT\tsample\(_name\)\{{0,1\}}/FORMAT\t{sample_name_str}/g' "
"| {fix_ambig} | vcfallelicprimitives -t DECOMPOSED --keep-geno | vcffixup - | vcfstreamsort "
"| {add_contig} {compress_cmd} > {tx_out_file}")
do.run(cl2.format(**locals()), "Finalising Scalpel variants", {})
return out_file
def run(align_bams, items, ref_file, assoc_files, region=None, out_file=None):
"""Run tumor only pisces calling
Handles bgzipping output file and fixing VCF sample naming to match BAM sample.
"""
paired = vcfutils.get_paired_bams(align_bams, items)
assert paired and not paired.normal_bam, (
"Pisces supports tumor-only variant calling: %s" %
(",".join([dd.get_sample_name(d) for d in items])))
vrs = bedutils.population_variant_regions(items)
target = shared.subset_variant_regions(vrs,
region,
out_file,
items=items,
do_merge=True)
min_af = float(dd.get_min_allele_fraction(paired.tumor_data)) / 100.0
if not utils.file_exists(out_file):
base_out_name = utils.splitext_plus(os.path.basename(
paired.tumor_bam))[0]
raw_file = "%s.vcf" % utils.splitext_plus(out_file)[0]
with file_transaction(paired.tumor_data, raw_file) as tx_out_file:
ref_dir = _prep_genome(os.path.dirname(tx_out_file),
paired.tumor_data)
out_dir = os.path.dirname(tx_out_file)
cores = dd.get_num_cores(paired.tumor_data)
cmd = (
"pisces --bampaths {paired.tumor_bam} --genomepaths {ref_dir} --intervalpaths {target} "
"--maxthreads {cores} --minvf {min_af} --ploidy somatic --gvcf false -o {out_dir}"
)
do.run(cmd.format(**locals()), "Pisces tumor-only somatic calling")
shutil.move(os.path.join(out_dir, "%s.vcf" % base_out_name),
tx_out_file)
vcfutils.bgzip_and_index(
raw_file,
paired.tumor_data["config"],
prep_cmd="sed 's#%s.bam#%s#' | %s" %
(base_out_name, dd.get_sample_name(paired.tumor_data),
vcfutils.add_contig_to_header_cl(
dd.get_ref_file(paired.tumor_data), out_file)))
return vcfutils.bgzip_and_index(out_file, paired.tumor_data["config"])
def run(align_bams, items, ref_file, assoc_files, region=None, out_file=None):
"""Run tumor only pisces calling
Handles bgzipping output file and fixing VCF sample naming to match BAM sample.
"""
paired = vcfutils.get_paired_bams(align_bams, items)
assert paired and not paired.normal_bam, ("Pisces supports tumor-only variant calling: %s" %
(",".join([dd.get_sample_name(d) for d in items])))
vrs = bedutils.population_variant_regions(items)
target = shared.subset_variant_regions(vrs, region,
out_file, items=items, do_merge=True)
min_af = float(dd.get_min_allele_fraction(paired.tumor_data)) / 100.0
if not utils.file_exists(out_file):
base_out_name = utils.splitext_plus(os.path.basename(paired.tumor_bam))[0]
raw_file = "%s.vcf" % utils.splitext_plus(out_file)[0]
with file_transaction(paired.tumor_data, raw_file) as tx_out_file:
ref_dir = _prep_genome(os.path.dirname(tx_out_file), paired.tumor_data)
out_dir = os.path.dirname(tx_out_file)
cores = dd.get_num_cores(paired.tumor_data)
emit_min_af = min_af / 10.0
cmd = ("pisces --bampaths {paired.tumor_bam} --genomepaths {ref_dir} --intervalpaths {target} "
"--maxthreads {cores} --minvf {emit_min_af} --vffilter {min_af} "
"--ploidy somatic --gvcf false -o {out_dir}")
# Recommended filtering for low frequency indels
# https://github.com/bcbio/bcbio-nextgen/commit/49d0cbb1f6dcbea629c63749e2f9813bd06dcee3#commitcomment-29765373
cmd += " -RMxNFilter 5,9,0.35"
# For low frequency UMI tagged variants, set higher variant thresholds
# https://github.com/Illumina/Pisces/issues/14#issuecomment-399756862
if min_af < (1.0 / 100.0):
cmd += " --minbasecallquality 30"
do.run(cmd.format(**locals()), "Pisces tumor-only somatic calling")
shutil.move(os.path.join(out_dir, "%s.vcf" % base_out_name),
tx_out_file)
vcfutils.bgzip_and_index(raw_file, paired.tumor_data["config"],
prep_cmd="sed 's#%s.bam#%s#' | %s" %
(base_out_name, dd.get_sample_name(paired.tumor_data),
vcfutils.add_contig_to_header_cl(dd.get_ref_file(paired.tumor_data), out_file)))
return vcfutils.bgzip_and_index(out_file, paired.tumor_data["config"])
def _produce_compatible_vcf(out_file, data):
"""Create a compatible VCF that downstream tools can deal with.
- htsjdk and thus GATK and Picard do not support VCF4.3:
https://github.com/broadinstitute/gatk/issues/2092
- Use octopus legacy format to avoid incompatibilities.
https://github.com/luntergroup/octopus#output-format
- Fixes `##contig` lines since octopus only writes contigs
used in the BED file region, causing incompatibilies with
GatherVcfs when merging
"""
base, ext = utils.splitext_plus(out_file)
legacy_file = "%s.legacy%s" % (base, ext)
final_file = "%s.vcf.gz" % base
cat_cmd = "zcat" if legacy_file.endswith(".gz") else "cat"
contig_cl = vcfutils.add_contig_to_header_cl(dd.get_ref_file(data),
out_file)
cmd = (
"{cat_cmd} {legacy_file} | sed 's/fileformat=VCFv4.3/fileformat=VCFv4.2/' | "
"{contig_cl} | bgzip -c > {final_file}")
do.run(cmd.format(**locals()),
"Produce compatible VCF output file from octopus")
return vcfutils.bgzip_and_index(out_file, data["config"])
def _run_vardict_paired(align_bams,
items,
ref_file,
assoc_files,
region=None,
out_file=None):
"""Detect variants with Vardict.
This is used for paired tumor / normal samples.
"""
config = items[0]["config"]
if out_file is None:
out_file = "%s-paired-variants.vcf.gz" % os.path.splitext(
align_bams[0])[0]
if not utils.file_exists(out_file):
with file_transaction(items[0], out_file) as tx_out_file:
target = shared.subset_variant_regions(dd.get_variant_regions(
items[0]),
region,
out_file,
do_merge=True)
paired = vcfutils.get_paired_bams(align_bams, items)
if not _is_bed_file(target):
vcfutils.write_empty_vcf(
tx_out_file,
config,
samples=[
x for x in [paired.tumor_name, paired.normal_name] if x
])
else:
if not paired.normal_bam:
ann_file = _run_vardict_caller(align_bams, items, ref_file,
assoc_files, region,
out_file)
return ann_file
vardict = get_vardict_command(items[0])
vcfstreamsort = config_utils.get_program(
"vcfstreamsort", config)
strandbias = "testsomatic.R"
var2vcf = "var2vcf_paired.pl"
compress_cmd = "| bgzip -c" if out_file.endswith("gz") else ""
freq = float(
utils.get_in(config, ("algorithm", "min_allele_fraction"),
10)) / 100.0
# merge bed file regions as amplicon VarDict is only supported in single sample mode
opts, var2vcf_opts = _vardict_options_from_config(
items, config, out_file, target)
fix_ambig_ref = vcfutils.fix_ambiguous_cl()
fix_ambig_alt = vcfutils.fix_ambiguous_cl(5)
remove_dup = vcfutils.remove_dup_cl()
if any("vardict_somatic_filter" in tz.get_in((
"config", "algorithm", "tools_off"), data, [])
for data in items):
somatic_filter = ""
freq_filter = ""
else:
var2vcf_opts += " -M " # this makes VarDict soft filter non-differential variants
somatic_filter = (
"| sed 's/\\\\.*Somatic\\\\/Somatic/' "
"| sed 's/REJECT,Description=\".*\">/REJECT,Description=\"Not Somatic via VarDict\">/' "
"""| %s -c 'from bcbio.variation import freebayes; """
"""freebayes.call_somatic("%s", "%s")' """ %
(sys.executable, paired.tumor_name,
paired.normal_name))
freq_filter = (
"| bcftools filter -m '+' -s 'REJECT' -e 'STATUS !~ \".*Somatic\"' 2> /dev/null "
"| %s -x 'bcbio.variation.vardict.depth_freq_filter(x, %s, \"%s\")'"
% (os.path.join(os.path.dirname(sys.executable), "py"),
0, dd.get_aligner(paired.tumor_data)))
jvm_opts = _get_jvm_opts(items[0], tx_out_file)
py_cl = os.path.join(utils.get_bcbio_bin(), "py")
setup = ("%s && unset JAVA_HOME &&" % utils.get_R_exports())
contig_cl = vcfutils.add_contig_to_header_cl(
ref_file, tx_out_file)
cmd = (
"{setup}{jvm_opts}{vardict} -G {ref_file} -f {freq} "
"-N {paired.tumor_name} -b \"{paired.tumor_bam}|{paired.normal_bam}\" {opts} "
"| {strandbias} "
"| {var2vcf} -P 0.9 -m 4.25 -f {freq} {var2vcf_opts} "
"-N \"{paired.tumor_name}|{paired.normal_name}\" "
"| {contig_cl} {freq_filter} "
"| bcftools filter -i 'QUAL >= 0' "
"{somatic_filter} | {fix_ambig_ref} | {fix_ambig_alt} | {remove_dup} | {vcfstreamsort} "
"{compress_cmd} > {tx_out_file}")
do.run(cmd.format(**locals()),
"Genotyping with VarDict: Inference", {})
return out_file
def _run_vardict_caller(align_bams,
items,
ref_file,
assoc_files,
region=None,
out_file=None):
"""Detect SNPs and indels with VarDict.
var2vcf_valid uses -A flag which reports all alleles and improves sensitivity:
https://github.com/AstraZeneca-NGS/VarDict/issues/35#issuecomment-276738191
"""
config = items[0]["config"]
if out_file is None:
out_file = "%s-variants.vcf.gz" % os.path.splitext(align_bams[0])[0]
if not utils.file_exists(out_file):
with file_transaction(items[0], out_file) as tx_out_file:
vrs = bedutils.population_variant_regions(items)
target = shared.subset_variant_regions(vrs,
region,
out_file,
items=items,
do_merge=False)
num_bams = len(align_bams)
sample_vcf_names = [
] # for individual sample names, given batch calling may be required
for bamfile, item in zip(align_bams, items):
# prepare commands
sample = dd.get_sample_name(item)
vardict = get_vardict_command(items[0])
strandbias = "teststrandbias.R"
var2vcf = "var2vcf_valid.pl"
opts, var2vcf_opts = _vardict_options_from_config(
items, config, out_file, target)
vcfstreamsort = config_utils.get_program(
"vcfstreamsort", config)
compress_cmd = "| bgzip -c" if tx_out_file.endswith(
"gz") else ""
freq = float(
utils.get_in(config, ("algorithm", "min_allele_fraction"),
10)) / 100.0
fix_ambig_ref = vcfutils.fix_ambiguous_cl()
fix_ambig_alt = vcfutils.fix_ambiguous_cl(5)
remove_dup = vcfutils.remove_dup_cl()
py_cl = os.path.join(utils.get_bcbio_bin(), "py")
jvm_opts = _get_jvm_opts(items[0], tx_out_file)
setup = ("%s && unset JAVA_HOME &&" % utils.get_R_exports())
contig_cl = vcfutils.add_contig_to_header_cl(
ref_file, tx_out_file)
cmd = (
"{setup}{jvm_opts}{vardict} -G {ref_file} -f {freq} "
"-N {sample} -b {bamfile} {opts} "
"| {strandbias}"
"| {var2vcf} -A -N {sample} -E -f {freq} {var2vcf_opts} "
"| {contig_cl} | bcftools filter -i 'QUAL >= 0' "
"| {fix_ambig_ref} | {fix_ambig_alt} | {remove_dup} | {vcfstreamsort} {compress_cmd}"
)
if num_bams > 1:
temp_file_prefix = out_file.replace(".gz", "").replace(
".vcf", "") + item["name"][1]
tmp_out = temp_file_prefix + ".temp.vcf"
tmp_out += ".gz" if out_file.endswith("gz") else ""
sample_vcf_names.append(tmp_out)
with file_transaction(item, tmp_out) as tx_tmp_file:
if not _is_bed_file(target):
vcfutils.write_empty_vcf(tx_tmp_file,
config,
samples=[sample])
else:
cmd += " > {tx_tmp_file}"
do.run(cmd.format(**locals()),
"Genotyping with VarDict: Inference", {})
else:
if not _is_bed_file(target):
vcfutils.write_empty_vcf(tx_out_file,
config,
samples=[sample])
else:
cmd += " > {tx_out_file}"
do.run(cmd.format(**locals()),
"Genotyping with VarDict: Inference", {})
if num_bams > 1:
# N.B. merge_variant_files wants region in 1-based end-inclusive
# coordinates. Thus use bamprep.region_to_gatk
vcfutils.merge_variant_files(
orig_files=sample_vcf_names,
out_file=tx_out_file,
ref_file=ref_file,
config=config,
region=bamprep.region_to_gatk(region))
return out_file
def _run_scalpel_paired(align_bams,
items,
ref_file,
assoc_files,
region=None,
out_file=None):
"""Detect indels with Scalpel.
This is used for paired tumor / normal samples.
"""
config = items[0]["config"]
if out_file is None:
out_file = "%s-paired-variants.vcf.gz" % os.path.splitext(
align_bams[0])[0]
if not utils.file_exists(out_file):
with file_transaction(config, out_file) as tx_out_file:
paired = get_paired_bams(align_bams, items)
if not paired.normal_bam:
ann_file = _run_scalpel_caller(align_bams, items, ref_file,
assoc_files, region, out_file)
return ann_file
vcfstreamsort = config_utils.get_program("vcfstreamsort", config)
perl_exports = utils.get_perl_exports(os.path.dirname(tx_out_file))
tmp_path = "%s-scalpel-work" % utils.splitext_plus(out_file)[0]
db_file = os.path.join(tmp_path, "main", "somatic.db")
if not os.path.exists(db_file + ".dir"):
if os.path.exists(tmp_path):
utils.remove_safe(tmp_path)
opts = " ".join(
_scalpel_options_from_config(items, config, out_file,
region, tmp_path))
opts += " --ref {}".format(ref_file)
opts += " --dir %s" % tmp_path
# caling
cl = (
"{perl_exports} && "
"scalpel-discovery --somatic {opts} --tumor {paired.tumor_bam} --normal {paired.normal_bam}"
)
do.run(cl.format(**locals()),
"Genotyping paired variants with Scalpel", {})
# filtering to adjust input parameters
bed_opts = " ".join(
_scalpel_bed_file_opts(items, config, out_file, region,
tmp_path))
use_defaults = True
if use_defaults:
scalpel_tmp_file = os.path.join(tmp_path,
"main/somatic.indel.vcf")
# Uses default filters but can tweak min-alt-count-tumor and min-phred-fisher
# to swap precision for sensitivity
else:
scalpel_tmp_file = os.path.join(
tmp_path, "main/somatic-indel-filter.vcf.gz")
with file_transaction(config,
scalpel_tmp_file) as tx_indel_file:
cmd = (
"{perl_exports} && "
"scalpel-export --somatic {bed_opts} --ref {ref_file} --db {db_file} "
"--min-alt-count-tumor 5 --min-phred-fisher 10 --min-vaf-tumor 0.1 "
"| bgzip -c > {tx_indel_file}")
do.run(cmd.format(**locals()),
"Scalpel somatic indel filter", {})
scalpel_tmp_file = bgzip_and_index(scalpel_tmp_file, config)
scalpel_tmp_file_common = bgzip_and_index(
os.path.join(tmp_path, "main/common.indel.vcf"), config)
compress_cmd = "| bgzip -c" if out_file.endswith("gz") else ""
bcftools_cmd_chi2 = get_scalpel_bcftools_filter_expression(
"chi2", config)
bcftools_cmd_common = get_scalpel_bcftools_filter_expression(
"reject", config)
fix_ambig = vcfutils.fix_ambiguous_cl()
add_contig = vcfutils.add_contig_to_header_cl(items[0])
cl2 = (
"vcfcat <({bcftools_cmd_chi2} {scalpel_tmp_file}) "
"<({bcftools_cmd_common} {scalpel_tmp_file_common}) | "
" {fix_ambig} | {vcfstreamsort} | {add_contig} {compress_cmd} > {tx_out_file}"
)
do.run(cl2.format(**locals()), "Finalising Scalpel variants", {})
return out_file
def _run_vardict_caller(align_bams, items, ref_file, assoc_files,
region=None, out_file=None):
"""Detect SNPs and indels with VarDict.
var2vcf_valid uses -A flag which reports all alleles and improves sensitivity:
https://github.com/AstraZeneca-NGS/VarDict/issues/35#issuecomment-276738191
"""
config = items[0]["config"]
if out_file is None:
out_file = "%s-variants.vcf.gz" % os.path.splitext(align_bams[0])[0]
if not utils.file_exists(out_file):
with file_transaction(items[0], out_file) as tx_out_file:
vrs = bedutils.population_variant_regions(items)
target = shared.subset_variant_regions(
vrs, region, out_file, items=items, do_merge=False)
num_bams = len(align_bams)
sample_vcf_names = [] # for individual sample names, given batch calling may be required
for bamfile, item in zip(align_bams, items):
# prepare commands
sample = dd.get_sample_name(item)
vardict = get_vardict_command(items[0])
opts, var2vcf_opts = _vardict_options_from_config(items, config, out_file, target)
vcfstreamsort = config_utils.get_program("vcfstreamsort", config)
compress_cmd = "| bgzip -c" if tx_out_file.endswith("gz") else ""
freq = float(utils.get_in(config, ("algorithm", "min_allele_fraction"), 10)) / 100.0
fix_ambig_ref = vcfutils.fix_ambiguous_cl()
fix_ambig_alt = vcfutils.fix_ambiguous_cl(5)
remove_dup = vcfutils.remove_dup_cl()
py_cl = os.path.join(utils.get_bcbio_bin(), "py")
jvm_opts = _get_jvm_opts(items[0], tx_out_file)
setup = ("%s && unset JAVA_HOME &&" % utils.get_R_exports())
contig_cl = vcfutils.add_contig_to_header_cl(ref_file, tx_out_file)
lowfreq_filter = _lowfreq_linear_filter(0, False)
cmd = ("{setup}{jvm_opts}{vardict} -G {ref_file} -f {freq} "
"-N {sample} -b {bamfile} {opts} "
"| teststrandbias.R "
"| var2vcf_valid.pl -A -N {sample} -E -f {freq} {var2vcf_opts} "
"| {contig_cl} | bcftools filter -i 'QUAL >= 0' | {lowfreq_filter} "
"| {fix_ambig_ref} | {fix_ambig_alt} | {remove_dup} | {vcfstreamsort} {compress_cmd}")
if num_bams > 1:
temp_file_prefix = out_file.replace(".gz", "").replace(".vcf", "") + item["name"][1]
tmp_out = temp_file_prefix + ".temp.vcf"
tmp_out += ".gz" if out_file.endswith("gz") else ""
sample_vcf_names.append(tmp_out)
with file_transaction(item, tmp_out) as tx_tmp_file:
if not _is_bed_file(target):
vcfutils.write_empty_vcf(tx_tmp_file, config, samples=[sample])
else:
cmd += " > {tx_tmp_file}"
do.run(cmd.format(**locals()), "Genotyping with VarDict: Inference", {})
else:
if not _is_bed_file(target):
vcfutils.write_empty_vcf(tx_out_file, config, samples=[sample])
else:
cmd += " > {tx_out_file}"
do.run(cmd.format(**locals()), "Genotyping with VarDict: Inference", {})
if num_bams > 1:
# N.B. merge_variant_files wants region in 1-based end-inclusive
# coordinates. Thus use bamprep.region_to_gatk
vcfutils.merge_variant_files(orig_files=sample_vcf_names,
out_file=tx_out_file, ref_file=ref_file,
config=config, region=bamprep.region_to_gatk(region))
return out_file
def _run_vardict_paired(align_bams, items, ref_file, assoc_files,
region=None, out_file=None):
"""Detect variants with Vardict.
This is used for paired tumor / normal samples.
"""
config = items[0]["config"]
if out_file is None:
out_file = "%s-paired-variants.vcf.gz" % os.path.splitext(align_bams[0])[0]
if not utils.file_exists(out_file):
with file_transaction(items[0], out_file) as tx_out_file:
vrs = bedutils.population_variant_regions(items)
target = shared.subset_variant_regions(vrs, region,
out_file, items=items, do_merge=True)
paired = vcfutils.get_paired_bams(align_bams, items)
if not _is_bed_file(target):
vcfutils.write_empty_vcf(tx_out_file, config,
samples=[x for x in [paired.tumor_name, paired.normal_name] if x])
else:
if not paired.normal_bam:
ann_file = _run_vardict_caller(align_bams, items, ref_file,
assoc_files, region, out_file)
return ann_file
vardict = get_vardict_command(items[0])
vcfstreamsort = config_utils.get_program("vcfstreamsort", config)
compress_cmd = "| bgzip -c" if out_file.endswith("gz") else ""
freq = float(utils.get_in(config, ("algorithm", "min_allele_fraction"), 10)) / 100.0
# merge bed file regions as amplicon VarDict is only supported in single sample mode
opts, var2vcf_opts = _vardict_options_from_config(items, config, out_file, target)
fix_ambig_ref = vcfutils.fix_ambiguous_cl()
fix_ambig_alt = vcfutils.fix_ambiguous_cl(5)
remove_dup = vcfutils.remove_dup_cl()
if any("vardict_somatic_filter" in tz.get_in(("config", "algorithm", "tools_off"), data, [])
for data in items):
somatic_filter = ""
freq_filter = ""
else:
var2vcf_opts += " -M " # this makes VarDict soft filter non-differential variants
somatic_filter = ("| sed 's/\\\\.*Somatic\\\\/Somatic/' "
"| sed 's/REJECT,Description=\".*\">/REJECT,Description=\"Not Somatic via VarDict\">/' "
"""| %s -c 'from bcbio.variation import freebayes; """
"""freebayes.call_somatic("%s", "%s")' """
% (sys.executable, paired.tumor_name, paired.normal_name))
freq_filter = ("| bcftools filter -m '+' -s 'REJECT' -e 'STATUS !~ \".*Somatic\"' 2> /dev/null "
"| %s -x 'bcbio.variation.vardict.add_db_germline_flag(x)' "
"| %s "
"| %s -x 'bcbio.variation.vardict.depth_freq_filter(x, %s, \"%s\")'" %
(os.path.join(os.path.dirname(sys.executable), "py"),
_lowfreq_linear_filter(0, True),
os.path.join(os.path.dirname(sys.executable), "py"),
0, bam.aligner_from_header(paired.tumor_bam)))
jvm_opts = _get_jvm_opts(items[0], tx_out_file)
py_cl = os.path.join(utils.get_bcbio_bin(), "py")
setup = ("%s && unset JAVA_HOME &&" % utils.get_R_exports())
contig_cl = vcfutils.add_contig_to_header_cl(ref_file, tx_out_file)
cmd = ("{setup}{jvm_opts}{vardict} -G {ref_file} -f {freq} "
"-N {paired.tumor_name} -b \"{paired.tumor_bam}|{paired.normal_bam}\" {opts} "
"| awk 'NF>=48' | testsomatic.R "
"| var2vcf_paired.pl -P 0.9 -m 4.25 -f {freq} {var2vcf_opts} "
"-N \"{paired.tumor_name}|{paired.normal_name}\" "
"| {contig_cl} {freq_filter} "
"| bcftools filter -i 'QUAL >= 0' "
"{somatic_filter} | {fix_ambig_ref} | {fix_ambig_alt} | {remove_dup} | {vcfstreamsort} "
"{compress_cmd} > {tx_out_file}")
do.run(cmd.format(**locals()), "Genotyping with VarDict: Inference", {})
return out_file
