def prep_recal(data):
"""Do pre-BQSR recalibration, calculation of recalibration tables.
"""
if dd.get_recalibrate(data) in [True, "gatk"]:
logger.info("Prepare BQSR tables with GATK: %s " %
str(dd.get_sample_name(data)))
dbsnp_file = tz.get_in(("genome_resources", "variation", "dbsnp"),
data)
if not dbsnp_file:
logger.info(
"Skipping GATK BaseRecalibrator because no VCF file of known variants was found."
)
return data
broad_runner = broad.runner_from_config(data["config"])
data["prep_recal"] = _gatk_base_recalibrator(
broad_runner, dd.get_align_bam(data), dd.get_ref_file(data),
dd.get_platform(data), dbsnp_file, dd.get_variant_regions(data),
data)
elif dd.get_recalibrate(data) == "sentieon":
logger.info("Prepare BQSR tables with sentieon: %s " %
str(dd.get_sample_name(data)))
data["prep_recal"] = sentieon.bqsr_table(data)
elif dd.get_recalibrate(data):
raise NotImplementedError("Unsupported recalibration type: %s" %
(dd.get_recalibrate(data)))
return data
def _do_run(paired):
"""Perform Battenberg caling with the paired dataset.
This purposely does not use a temporary directory for the output
since Battenberg does smart restarts.
"""
work_dir = _sv_workdir(paired.tumor_data)
ignore_file = os.path.join(work_dir, "ignore_chromosomes.txt")
out = _get_battenberg_out(paired, work_dir)
if len(_missing_files(out)) > 0:
ref_file = dd.get_ref_file(paired.tumor_data)
bat_datadir = os.path.normpath(os.path.join(os.path.dirname(ref_file), os.pardir, "battenberg"))
ignore_file = _make_ignore_file(work_dir, ref_file, os.path.join(bat_datadir, "impute", "impute_info.txt"),
ignore_file)
local_sitelib = os.path.join(install.get_defaults().get("tooldir", "/usr/local"),
"lib", "R", "site-library")
perllib = os.path.join(install.get_defaults().get("tooldir", "/usr/local"),
"lib", "perl5")
tumor_bam = paired.tumor_bam
normal_bam = paired.normal_bam
platform = dd.get_platform(paired.tumor_data)
genome_build = paired.tumor_data["genome_build"]
# scale cores to avoid over-using memory during imputation
cores = max(1, int(dd.get_num_cores(paired.tumor_data) * 0.5))
cmd = ("export R_LIBS_USER={local_sitelib} && "
"export PERL5LIB={perllib}:$PERL5LIB "
"battenberg.pl -t {cores} -o {work_dir} -r {ref_file}.fai "
"-tb {tumor_bam} -nb {normal_bam} -e {bat_datadir}/impute/impute_info.txt "
"-u {bat_datadir}/1000genomesloci -c {bat_datadir}/probloci.txt "
"-ig {ignore_file} "
"-assembly {genome_build} -species Human -platform {platform}")
do.run(cmd.format(**locals()), "Battenberg CNV calling")
assert len(_missing_files(out)) == 0, "Missing Battenberg output: %s" % _missing_files(out)
out["ignore"] = ignore_file
return out
def _do_run(paired):
"""Perform Battenberg caling with the paired dataset.
This purposely does not use a temporary directory for the output
since Battenberg does smart restarts.
"""
work_dir = _sv_workdir(paired.tumor_data)
out = _get_battenberg_out(paired, work_dir)
ignore_file = os.path.join(work_dir, "ignore_chromosomes.txt")
if len(_missing_files(out)) > 0:
ref_file = dd.get_ref_file(paired.tumor_data)
bat_datadir = os.path.normpath(
os.path.join(os.path.dirname(ref_file), os.pardir, "battenberg"))
ignore_file, gl_file = _make_ignore_file(
work_dir, ref_file, ignore_file,
os.path.join(bat_datadir, "impute", "impute_info.txt"))
local_sitelib = os.path.join(
install.get_defaults().get("tooldir", "/usr/local"), "lib", "R",
"site-library")
tumor_bam = paired.tumor_bam
normal_bam = paired.normal_bam
platform = dd.get_platform(paired.tumor_data)
genome_build = paired.tumor_data["genome_build"]
# scale cores to avoid over-using memory during imputation
cores = max(1, int(dd.get_num_cores(paired.tumor_data) * 0.5))
gender = {
"male": "XY",
"female": "XX",
"unknown": "L"
}.get(population.get_gender(paired.tumor_data))
if gender == "L":
gender_str = "-ge %s -gl %s" % (gender, gl_file)
else:
gender_str = "-ge %s" % (gender)
r_export_cmd = "unset R_HOME && export PATH=%s:$PATH && " % os.path.dirname(
utils.Rscript_cmd())
cmd = (
"export R_LIBS_USER={local_sitelib} && {r_export_cmd}"
"battenberg.pl -t {cores} -o {work_dir} -r {ref_file}.fai "
"-tb {tumor_bam} -nb {normal_bam} -e {bat_datadir}/impute/impute_info.txt "
"-u {bat_datadir}/1000genomesloci -c {bat_datadir}/probloci.txt "
"-ig {ignore_file} {gender_str} "
"-assembly {genome_build} -species Human -platform {platform}")
do.run(cmd.format(**locals()), "Battenberg CNV calling")
assert len(_missing_files(
out)) == 0, "Missing Battenberg output: %s" % _missing_files(out)
out["plot"] = _get_battenberg_out_plots(paired, work_dir)
out["ignore"] = ignore_file
return out
def prep_recal(data):
"""Do pre-BQSR recalibration, calculation of recalibration tables.
"""
if dd.get_recalibrate(data) in [True, "gatk"]:
logger.info("Prepare BQSR tables with GATK: %s " % str(dd.get_sample_name(data)))
dbsnp_file = tz.get_in(("genome_resources", "variation", "dbsnp"), data)
if not dbsnp_file:
logger.info("Skipping GATK BaseRecalibrator because no VCF file of known variants was found.")
return data
broad_runner = broad.runner_from_config(data["config"])
data["prep_recal"] = _gatk_base_recalibrator(broad_runner, dd.get_align_bam(data),
dd.get_ref_file(data), dd.get_platform(data),
dbsnp_file, dd.get_variant_regions(data), data)
elif dd.get_recalibrate(data) == "sentieon":
logger.info("Prepare BQSR tables with sentieon: %s " % str(dd.get_sample_name(data)))
data["prep_recal"] = sentieon.bqsr_table(data)
elif dd.get_recalibrate(data):
raise NotImplementedError("Unsupported recalibration type: %s" % (dd.get_recalibrate(data)))
return data
def _do_run(paired):
"""Perform Battenberg caling with the paired dataset.
This purposely does not use a temporary directory for the output
since Battenberg does smart restarts.
"""
work_dir = _sv_workdir(paired.tumor_data)
out = _get_battenberg_out(paired, work_dir)
ignore_file = os.path.join(work_dir, "ignore_chromosomes.txt")
if len(_missing_files(out)) > 0:
ref_file = dd.get_ref_file(paired.tumor_data)
bat_datadir = os.path.normpath(os.path.join(os.path.dirname(ref_file), os.pardir, "battenberg"))
ignore_file, gl_file = _make_ignore_file(work_dir, ref_file, ignore_file,
os.path.join(bat_datadir, "impute", "impute_info.txt"))
local_sitelib = os.path.join(install.get_defaults().get("tooldir", "/usr/local"),
"lib", "R", "site-library")
tumor_bam = paired.tumor_bam
normal_bam = paired.normal_bam
platform = dd.get_platform(paired.tumor_data)
genome_build = paired.tumor_data["genome_build"]
# scale cores to avoid over-using memory during imputation
cores = max(1, int(dd.get_num_cores(paired.tumor_data) * 0.5))
gender = {"male": "XY", "female": "XX", "unknown": "L"}.get(population.get_gender(paired.tumor_data))
if gender == "L":
gender_str = "-ge %s -gl %s" % (gender, gl_file)
else:
gender_str = "-ge %s" % (gender)
r_export_cmd = "unset R_HOME && export PATH=%s:$PATH && " % os.path.dirname(utils.Rscript_cmd())
cmd = ("export R_LIBS_USER={local_sitelib} && {r_export_cmd}"
"battenberg.pl -t {cores} -o {work_dir} -r {ref_file}.fai "
"-tb {tumor_bam} -nb {normal_bam} -e {bat_datadir}/impute/impute_info.txt "
"-u {bat_datadir}/1000genomesloci -c {bat_datadir}/probloci.txt "
"-ig {ignore_file} {gender_str} "
"-assembly {genome_build} -species Human -platform {platform}")
do.run(cmd.format(**locals()), "Battenberg CNV calling")
assert len(_missing_files(out)) == 0, "Missing Battenberg output: %s" % _missing_files(out)
out["plot"] = _get_battenberg_out_plots(paired, work_dir)
out["ignore"] = ignore_file
return out