def __init__(self,
config,
reference,
input_bam,
output_bam,
tags,
min_mapq=0,
filter_unmapped=False,
dependencies=()):
flt_params = AtomicCmdBuilder(("samtools", "view", "-bu"),
IN_BAM=input_bam,
OUT_STDOUT=AtomicCmd.PIPE)
if min_mapq:
flt_params.set_option("-q", min_mapq, sep="")
if filter_unmapped:
flt_params.set_option("-F", "0x4", sep="")
flt_params.add_value("%(IN_BAM)s")
jar_params = picard.picard_command(config, "AddOrReplaceReadGroups")
jar_params.set_option("INPUT", "/dev/stdin", sep="=")
# Output is written to a named pipe, since the JVM may, in some cases,
# emit warning messages to stdout, resulting in a malformed BAM.
jar_params.set_option("OUTPUT", "%(TEMP_OUT_BAM)s", sep="=")
jar_params.set_option("COMPRESSION_LEVEL", "0", sep="=")
# Ensure that the BAM is sorted; this is required by the pipeline, and
# needs to be done before calling calmd (avoiding pathologic runtimes).
jar_params.set_option("SORT_ORDER", "coordinate", sep="=")
# All tags are overwritten; ID is set since the default (e.g. '1')
# causes problems with pysam due to type inference (is read as a length
# 1 string, but written as a character).
for tag in ("ID", "SM", "LB", "PU", "PL"):
jar_params.set_option(tag, tags[tag], sep="=")
jar_params.set_kwargs(IN_STDIN=flt_params, TEMP_OUT_BAM="bam.pipe")
calmd = AtomicCmdBuilder(
["samtools", "calmd", "-b", "%(TEMP_IN_BAM)s", "%(IN_REF)s"],
IN_REF=reference,
TEMP_IN_BAM="bam.pipe",
OUT_STDOUT=output_bam)
commands = [cmd.finalize() for cmd in (flt_params, jar_params, calmd)]
description = "<Cleanup BAM: %s -> '%s'>" \
% (input_bam, output_bam)
PicardNode.__init__(self,
command=ParallelCmds(commands),
description=description,
dependencies=dependencies)
def __init__(self, config, reference, input_bam, output_bam, tags,
min_mapq=0, filter_unmapped=False, dependencies=()):
flt_params = AtomicCmdBuilder(("samtools", "view", "-bu"),
IN_BAM=input_bam,
OUT_STDOUT=AtomicCmd.PIPE)
if min_mapq:
flt_params.set_option("-q", min_mapq, sep="")
if filter_unmapped:
flt_params.set_option("-F", "0x4", sep="")
flt_params.add_value("%(IN_BAM)s")
jar_params = picard.picard_command(config, "AddOrReplaceReadGroups")
jar_params.set_option("INPUT", "/dev/stdin", sep="=")
# Output is written to a named pipe, since the JVM may, in some cases,
# emit warning messages to stdout, resulting in a malformed BAM.
jar_params.set_option("OUTPUT", "%(TEMP_OUT_BAM)s", sep="=")
jar_params.set_option("COMPRESSION_LEVEL", "0", sep="=")
# Ensure that the BAM is sorted; this is required by the pipeline, and
# needs to be done before calling calmd (avoiding pathologic runtimes).
jar_params.set_option("SORT_ORDER", "coordinate", sep="=")
# All tags are overwritten; ID is set since the default (e.g. '1')
# causes problems with pysam due to type inference (is read as a length
# 1 string, but written as a character).
for tag in ("ID", "SM", "LB", "PU", "PL"):
jar_params.set_option(tag, tags[tag], sep="=")
jar_params.set_kwargs(IN_STDIN=flt_params,
TEMP_OUT_BAM="bam.pipe")
calmd = AtomicCmdBuilder(["samtools", "calmd", "-b",
"%(TEMP_IN_BAM)s", "%(IN_REF)s"],
IN_REF=reference,
TEMP_IN_BAM="bam.pipe",
OUT_STDOUT=output_bam)
commands = [cmd.finalize() for cmd in (flt_params, jar_params, calmd)]
description = "<Cleanup BAM: %s -> '%s'>" \
% (input_bam, output_bam)
PicardNode.__init__(self,
command=ParallelCmds(commands),
description=description,
dependencies=dependencies)
def __init__(self,
config,
reference,
input_bam,
output_bam,
tags,
min_mapq=0,
filter_unmapped=False,
dependencies=()):
call = ["samtools", "view", "-bu"]
if min_mapq > 0:
call.append("-q%i" % min_mapq)
if filter_unmapped:
call.append("-F0x4")
call.append("%(IN_BAM)s")
flt = AtomicCmd(call, IN_BAM=input_bam, OUT_STDOUT=AtomicCmd.PIPE)
jar_file = os.path.join(config.jar_root, "AddOrReplaceReadGroups.jar")
params = AtomicJavaCmdBuilder(jar=jar_file,
jre_options=config.jre_options)
params.set_option("INPUT", "/dev/stdin", sep="=")
params.set_option("OUTPUT", "%(TEMP_OUT_BAM)s", sep="=")
params.set_option("COMPRESSION_LEVEL", "0", sep="=")
params.set_option("SORT_ORDER", "coordinate", sep="=")
for tag in ("SM", "LB", "PU", "PL"):
params.set_option(tag, tags[tag], sep="=")
params.set_kwargs(IN_STDIN=flt, TEMP_OUT_BAM="bam.pipe")
annotate = params.finalize()
calmd = AtomicCmd(
["samtools", "calmd", "-b", "%(TEMP_IN_BAM)s", "%(IN_REF)s"],
IN_REF=reference,
TEMP_IN_BAM="bam.pipe",
OUT_STDOUT=output_bam)
description = "<Cleanup BAM: %s -> '%s'>" \
% (input_bam, output_bam)
PicardNode.__init__(self,
command=ParallelCmds([flt, annotate, calmd]),
description=description,
dependencies=dependencies)
def __init__(self, config, reference, input_bam, output_bam, tags,
min_mapq=0, filter_unmapped=False, dependencies=()):
call = ["samtools", "view", "-bu"]
if min_mapq > 0:
call.append("-q%i" % min_mapq)
if filter_unmapped:
call.append("-F0x4")
call.append("%(IN_BAM)s")
flt = AtomicCmd(call,
IN_BAM=input_bam,
OUT_STDOUT=AtomicCmd.PIPE)
jar_file = os.path.join(config.jar_root, "AddOrReplaceReadGroups.jar")
params = AtomicJavaCmdBuilder(jar=jar_file,
jre_options=config.jre_options)
params.set_option("INPUT", "/dev/stdin", sep="=")
params.set_option("OUTPUT", "%(TEMP_OUT_BAM)s", sep="=")
params.set_option("COMPRESSION_LEVEL", "0", sep="=")
params.set_option("SORT_ORDER", "coordinate", sep="=")
for tag in ("SM", "LB", "PU", "PL"):
params.set_option(tag, tags[tag], sep="=")
params.set_kwargs(IN_STDIN=flt,
TEMP_OUT_BAM="bam.pipe")
annotate = params.finalize()
calmd = AtomicCmd(["samtools", "calmd", "-b",
"%(TEMP_IN_BAM)s", "%(IN_REF)s"],
IN_REF=reference,
TEMP_IN_BAM="bam.pipe",
OUT_STDOUT=output_bam)
description = "<Cleanup BAM: %s -> '%s'>" \
% (input_bam, output_bam)
PicardNode.__init__(self,
command=ParallelCmds([flt, annotate, calmd]),
description=description,
dependencies=dependencies)
def _setup(self, config, temp_root):
PicardNode._setup(self, config, temp_root)
os.mkfifo(os.path.join(temp_root, "bam.pipe"))
def _setup(self, config, temp_root):
PicardNode._setup(self, config, temp_root)
os.mkfifo(os.path.join(temp_root, "bam.pipe"))
