def _square_batch_bcbio_variation(data, region, bam_files, vrn_files, out_file,
todo="square"):
"""Run squaring or merging analysis using bcbio.variation.recall.
"""
ref_file = tz.get_in(("reference", "fasta", "base"), data)
cores = tz.get_in(("config", "algorithm", "num_cores"), data, 1)
resources = config_utils.get_resources("bcbio-variation-recall", data["config"])
# adjust memory by cores but leave room for run program memory
memcores = int(math.ceil(float(cores) / 5.0))
jvm_opts = config_utils.adjust_opts(resources.get("jvm_opts", ["-Xms250m", "-Xmx2g"]),
{"algorithm": {"memory_adjust": {"direction": "increase",
"magnitude": memcores}}})
# Write unique VCFs and BAMs to input file
input_file = "%s-inputs.txt" % os.path.splitext(out_file)[0]
with open(input_file, "w") as out_handle:
out_handle.write("\n".join(sorted(list(set(vrn_files)))) + "\n")
if todo == "square":
out_handle.write("\n".join(sorted(list(set(bam_files)))) + "\n")
variantcaller = tz.get_in(("config", "algorithm", "jointcaller"), data).replace("-joint", "")
cmd = ["bcbio-variation-recall", todo] + jvm_opts + broad.get_default_jvm_opts() + \
["-c", cores, "-r", bamprep.region_to_gatk(region)]
if todo == "square":
cmd += ["--caller", variantcaller]
cmd += [out_file, ref_file, input_file]
bcbio_env = utils.get_bcbio_env()
cmd = " ".join(str(x) for x in cmd)
do.run(cmd, "%s in region: %s" % (cmd, bamprep.region_to_gatk(region)), env=bcbio_env)
return out_file
def _square_batch_bcbio_variation(data, region, bam_files, vrn_files, out_file,
todo="square"):
"""Run squaring or merging analysis using bcbio.variation.recall.
"""
ref_file = tz.get_in(("reference", "fasta", "base"), data)
cores = tz.get_in(("config", "algorithm", "num_cores"), data, 1)
resources = config_utils.get_resources("bcbio-variation-recall", data["config"])
# adjust memory by cores but leave room for run program memory
memcores = int(math.ceil(float(cores) / 5.0))
jvm_opts = config_utils.adjust_opts(resources.get("jvm_opts", ["-Xms250m", "-Xmx2g"]),
{"algorithm": {"memory_adjust": {"direction": "increase",
"magnitude": memcores}}})
# Write unique VCFs and BAMs to input file
input_file = "%s-inputs.txt" % os.path.splitext(out_file)[0]
with open(input_file, "w") as out_handle:
out_handle.write("\n".join(sorted(list(set(vrn_files)))) + "\n")
if todo == "square":
out_handle.write("\n".join(sorted(list(set(bam_files)))) + "\n")
variantcaller = tz.get_in(("config", "algorithm", "jointcaller"), data).replace("-joint", "")
cmd = ["bcbio-variation-recall", todo] + jvm_opts + broad.get_default_jvm_opts() + \
["-c", cores, "-r", bamprep.region_to_gatk(region)]
if todo == "square":
cmd += ["--caller", variantcaller]
cmd += [out_file, ref_file, input_file]
do.run(cmd, "%s in region: %s" % (cmd, bamprep.region_to_gatk(region)))
return out_file
def _do_smash_calldiff(truth_vcf, eval_vcf, ref_file, out_dir, config):
resources = config_utils.get_resources("smash", config)
jvm_opts = resources.get("jvm_opts", ["-Xms250m", "-Xmx2g"])
cmd = ["smash"] + jvm_opts + broad.get_default_jvm_opts() + \
["--lhs_vcf", truth_vcf, "--rhs_vcf", eval_vcf,
"--reference_fasta", ref_file, "--presorted"]
do.run(cmd, "Compare files with SMaSH calldiff")
def _do_smash_calldiff(truth_vcf, eval_vcf, ref_file, out_dir, config):
resources = config_utils.get_resources("smash", config)
jvm_opts = resources.get("jvm_opts", ["-Xms250m", "-Xmx2g"])
cmd = ["smash"] + jvm_opts + broad.get_default_jvm_opts() + \
["--lhs_vcf", truth_vcf, "--rhs_vcf", eval_vcf,
"--reference_fasta", ref_file, "--presorted"]
do.run(cmd, "Compare files with SMaSH calldiff")
def bcbio_variation_comparison(config_file, base_dir, data):
"""Run a variant comparison using the bcbio.variation toolkit, given an input configuration.
"""
tmp_dir = utils.safe_makedir(os.path.join(base_dir, "tmp"))
resources = config_utils.get_resources("bcbio_variation", data["config"])
jvm_opts = resources.get("jvm_opts", ["-Xms750m", "-Xmx2g"])
cmd = ["bcbio-variation"] + jvm_opts + broad.get_default_jvm_opts(tmp_dir) + \
["variant-compare", config_file]
do.run(cmd, "Comparing variant calls using bcbio.variation", data)
def bcbio_variation_comparison(config_file, base_dir, data):
"""Run a variant comparison using the bcbio.variation toolkit, given an input configuration.
"""
tmp_dir = utils.safe_makedir(os.path.join(base_dir, "tmp"))
resources = config_utils.get_resources("bcbio_variation", data["config"])
jvm_opts = resources.get("jvm_opts", ["-Xms750m", "-Xmx2g"])
cmd = ["bcbio-variation"] + jvm_opts + broad.get_default_jvm_opts(tmp_dir) + \
["variant-compare", config_file]
do.run(cmd, "Comparing variant calls using bcbio.variation", data)
def _get_jvm_opts(data, out_file):
"""Retrieve JVM options when running the Java version of VarDict.
"""
if get_vardict_command(data) == "vardict-java":
resources = config_utils.get_resources("vardict", data["config"])
jvm_opts = resources.get("jvm_opts", ["-Xms750m", "-Xmx4g"])
jvm_opts += broad.get_default_jvm_opts(os.path.dirname(out_file))
return "export VAR_DICT_OPTS='%s' && " % " ".join(jvm_opts)
else:
return ""
def _get_jvm_opts(data, out_file):
"""Retrieve JVM options when running the Java version of VarDict.
"""
if dd.get_variantcaller(data).endswith("-java"):
resources = config_utils.get_resources("vardict", data["config"])
jvm_opts = resources.get("jvm_opts", ["-Xms750m", "-Xmx4g"])
jvm_opts += broad.get_default_jvm_opts(os.path.dirname(out_file))
return "export VAR_DICT_OPTS='%s' && " % " ".join(jvm_opts)
else:
return ""
def _get_jvm_opts(out_file, data):
"""Retrieve Java options, adjusting memory for available cores.
"""
resources = config_utils.get_resources("purple", data["config"])
jvm_opts = resources.get("jvm_opts", ["-Xms750m", "-Xmx3500m"])
jvm_opts = config_utils.adjust_opts(jvm_opts, {"algorithm": {"memory_adjust":
{"direction": "increase",
"maximum": "30000M",
"magnitude": dd.get_cores(data)}}})
jvm_opts += broad.get_default_jvm_opts(os.path.dirname(out_file))
return jvm_opts
def _get_fgbio_jvm_opts(data, tmpdir, scale_factor=None):
cores, mem = _get_cores_memory(data)
resources = config_utils.get_resources("fgbio", data["config"])
jvm_opts = resources.get("jvm_opts", ["-Xms750m", "-Xmx4g"])
if scale_factor and cores > scale_factor:
jvm_opts = config_utils.adjust_opts(jvm_opts, {"algorithm": {"memory_adjust":
{"direction": "increase",
"magnitude": cores // scale_factor}}})
jvm_opts += broad.get_default_jvm_opts()
jvm_opts = " ".join(jvm_opts)
return jvm_opts + " --tmp-dir %s" % tmpdir
def _get_fgbio_jvm_opts(data, tmpdir, scale_factor=None):
cores, mem = _get_cores_memory(data)
resources = config_utils.get_resources("fgbio", data["config"])
jvm_opts = resources.get("jvm_opts", ["-Xms750m", "-Xmx4g"])
if scale_factor and cores > scale_factor:
jvm_opts = config_utils.adjust_opts(jvm_opts, {"algorithm": {"memory_adjust":
{"direction": "increase",
"magnitude": cores // scale_factor}}})
jvm_opts += broad.get_default_jvm_opts()
jvm_opts = " ".join(jvm_opts)
return jvm_opts + " --tmp-dir %s" % tmpdir
def _get_varscan_opts(config, tmp_dir):
"""Retrieve common options for running VarScan.
Handles jvm_opts, setting user and country to English to avoid issues
with different locales producing non-compliant VCF.
"""
resources = config_utils.get_resources("varscan", config)
jvm_opts = resources.get("jvm_opts", ["-Xmx750m", "-Xmx2g"])
jvm_opts = config_utils.adjust_opts(jvm_opts,
{"algorithm": {"memory_adjust":
{"magnitude": 1.1, "direction": "decrease"}}})
jvm_opts += ["-Duser.language=en", "-Duser.country=US"]
jvm_opts += broad.get_default_jvm_opts(tmp_dir)
return " ".join(jvm_opts)
def _get_jvm_opts(out_file, data):
"""Retrieve Java options, adjusting memory for available cores.
"""
resources = config_utils.get_resources("purple", data["config"])
jvm_opts = resources.get("jvm_opts", ["-Xms750m", "-Xmx3500m"])
jvm_opts = config_utils.adjust_opts(
jvm_opts, {
"algorithm": {
"memory_adjust": {
"direction": "increase",
"maximum": "30000M",
"magnitude": dd.get_cores(data)
}
}
})
jvm_opts += broad.get_default_jvm_opts(os.path.dirname(out_file))
return jvm_opts
