def _picard_rna_metrics(readset):
"""
Computes a series of quality control metrics using both CollectRnaSeqMetrics and CollectAlignmentSummaryMetrics
functions metrics are collected using [Picard](http://broadinstitute.github.io/picard/).
"""
jobs = []
sample = readset.sample
alignment_file = readset.bam + ".bam"
output_directory = os.path.dirname(alignment_file)
job = picard.collect_multiple_metrics(
alignment_file,
readset.bam + ".metrics",
reference_sequence=readset.reference_file)
job.name = "picard_collect_multiple_metrics." + readset.name + ".met" + "." + readset.run + "." + readset.lane
jobs.append(job)
if len(readset.annotation_files) > 2 and os.path.isfile(
readset.annotation_files[2]):
job = picard.collect_rna_metrics(
alignment_file, os.path.join(output_directory, sample.name),
readset.annotation_files[2], readset.reference_file)
job.name = "picard_rna_metrics." + readset.name + ".rmet" + "." + readset.run + "." + readset.lane
jobs.append(job)
return jobs
def _picard_rna_metrics(readset):
"""
Computes a series of quality control metrics using both CollectRnaSeqMetrics and CollectAlignmentSummaryMetrics
functions metrics are collected using [Picard](http://broadinstitute.github.io/picard/).
"""
jobs = []
sample = readset.sample
alignment_file = readset.bam + ".bam"
output_directory = os.path.dirname(alignment_file)
job = picard.collect_multiple_metrics(alignment_file, readset.bam + ".metrics",
reference_sequence=readset.reference_file)
job.name = "picard_collect_multiple_metrics." + readset.name + ".met" + "." + readset.run + "." + readset.lane
jobs.append(job)
if len(readset.annotation_files) > 2 and os.path.isfile(readset.annotation_files[2]):
job = picard.collect_rna_metrics(alignment_file,
os.path.join(output_directory, sample.name),
readset.annotation_files[2],
readset.reference_file)
job.name = "picard_rna_metrics." + readset.name + ".rmet" + "." + readset.run + "." + readset.lane
jobs.append(job)
return jobs
def get_metrics_jobs(self, readset):
jobs = []
input_file_prefix = readset.bam + '.'
input = input_file_prefix + "bam"
job = picard.collect_multiple_metrics(input, input_file_prefix + "metrics",
reference_sequence=readset.reference_file)
job.name = "picard_collect_multiple_metrics." + readset.name + ".met" + "." + readset.run + "." + readset.lane
jobs.append(job)
if readset.beds:
coverage_bed = readset.beds[0]
full_coverage_bed = (self.output_dir + os.sep + coverage_bed)
else:
coverage_bed = None
full_coverage_bed = None
if coverage_bed:
if (not os.path.exists(full_coverage_bed)) and \
(coverage_bed not in BwaRunProcessingAligner.downloaded_bed_files):
# Download the bed file
command = config.param('DEFAULT', 'fetch_bed_file_command').format(
output_directory=self.output_dir,
filename=coverage_bed
)
job = Job([], [full_coverage_bed], command=command, name="bed_download." + coverage_bed)
BwaRunProcessingAligner.downloaded_bed_files.append(coverage_bed)
jobs.append(job)
interval_list = re.sub("\.[^.]+$", ".interval_list", coverage_bed)
if interval_list not in BwaRunProcessingAligner.created_interval_lists:
# Create one job to generate the interval list from the bed file
ref_dict = os.path.splitext(readset.reference_file)[0] + '.dict'
job = tools.bed2interval_list(ref_dict, full_coverage_bed, interval_list)
job.name = "interval_list." + coverage_bed
BwaRunProcessingAligner.created_interval_lists.append(interval_list)
jobs.append(job)
job = picard.calculate_hs_metrics(input_file_prefix + "bam", input_file_prefix + "metrics.onTarget.txt",
interval_list, reference_sequence=readset.reference_file)
job.name = "picard_calculate_hs_metrics." + readset.name + ".hs" + "." + readset.run + "." + readset.lane
jobs.append(job)
jobs.extend(self.verify_bam_id(readset, full_coverage_bed))
job = bvatools.depth_of_coverage(
input,
input_file_prefix + "metrics.targetCoverage.txt",
full_coverage_bed,
other_options=config.param('bvatools_depth_of_coverage', 'other_options', required=False),
reference_genome=readset.reference_file
)
job.name = "bvatools_depth_of_coverage." + readset.name + ".doc" + "." + readset.run + "." + readset.lane
jobs.append(job)
return jobs
def metrics(self):
jobs = []
for sample in self.samples:
file_prefix = os.path.join("alignment", sample.name, sample.name + ".sorted.dup.")
input = file_prefix + "bam"
job = picard.collect_multiple_metrics(input, file_prefix + "all.metrics")
job.name = "picard_collect_multiple_metrics." + sample.name
jobs.append(job)
# Compute genome coverage
job = gatk.depth_of_coverage(input, file_prefix + "all.coverage")
job.name = "gatk_depth_of_coverage.genome." + sample.name
jobs.append(job)
job = igvtools.compute_tdf(input, input + ".tdf")
job.name = "igvtools_compute_tdf." + sample.name
jobs.append(job)
return jobs
def picard_rna_metrics(self):
"""
Computes a series of quality control metrics using both CollectRnaSeqMetrics and CollectAlignmentSummaryMetrics functions
metrics are collected using [Picard](http://broadinstitute.github.io/picard/).
"""
jobs = []
reference_file = config.param('picard_rna_metrics', 'genome_fasta', type='filepath')
for sample in self.samples:
alignment_file = os.path.join("alignment", sample.name, sample.name + ".sorted.mdup.bam")
output_directory = os.path.join("metrics", sample.name)
job = concat_jobs([
Job(command="mkdir -p " + output_directory, removable_files=[output_directory]),
picard.collect_multiple_metrics(alignment_file, os.path.join(output_directory,sample.name),reference_file),
picard.collect_rna_metrics(alignment_file, os.path.join(output_directory,sample.name+".picard_rna_metrics"))
],name="picard_rna_metrics."+ sample.name)
jobs.append(job)
return jobs
def picard_rna_metrics(self):
"""
Computes a series of quality control metrics using both CollectRnaSeqMetrics and CollectAlignmentSummaryMetrics functions
metrics are collected using [Picard](http://broadinstitute.github.io/picard/).
"""
jobs = []
reference_file = config.param('picard_rna_metrics', 'genome_fasta', type='filepath')
for sample in self.samples:
alignment_file = os.path.join("alignment", sample.name, sample.name + ".sorted.mdup.bam")
output_directory = os.path.join("metrics", sample.name)
job = concat_jobs([
Job(command="mkdir -p " + output_directory, removable_files=[output_directory]),
picard.collect_multiple_metrics(alignment_file, os.path.join(output_directory,sample.name),reference_file),
picard.collect_rna_metrics(alignment_file, os.path.join(output_directory,sample.name+".picard_rna_metrics"))
],name="picard_rna_metrics."+ sample.name)
jobs.append(job)
return jobs
def metrics(self):
"""
Compute metrics and generate coverage tracks per sample. Multiple metrics are computed at this stage:
Number of raw reads, Number of filtered reads, Number of aligned reads, Number of duplicate reads,
Median, mean and standard deviation of insert sizes of reads after alignment, percentage of bases
covered at X reads (%_bases_above_50 means the % of exons bases which have at least 50 reads)
whole genome or targeted percentage of bases covered at X reads (%_bases_above_50 means the % of exons
bases which have at least 50 reads). A TDF (.tdf) coverage track is also generated at this step
for easy visualization of coverage in the IGV browser.
"""
jobs = []
for sample in self.samples:
input_file_prefix = os.path.join(
"alignment", sample.name, sample.name + ".matefixed.sorted.")
input = input_file_prefix + "bam"
job = picard.collect_multiple_metrics(
input, input_file_prefix + "all.metrics")
job.name = "picard_collect_multiple_metrics." + sample.name
job.samples = [sample]
jobs.append(job)
# Compute genome or target coverage with BVATools
job = bvatools.depth_of_coverage(
input,
input_file_prefix + "coverage.tsv",
bvatools.resolve_readset_coverage_bed(sample.readsets[0]),
other_options=config.param('bvatools_depth_of_coverage',
'other_options',
required=False))
job.name = "bvatools_depth_of_coverage." + sample.name
job.samples = [sample]
jobs.append(job)
job = igvtools.compute_tdf(input, input + ".tdf")
job.name = "igvtools_compute_tdf." + sample.name
job.samples = [sample]
jobs.append(job)
return jobs
def metrics(self):
"""
Compute metrics and generate coverage tracks per sample. Multiple metrics are computed at this stage:
Number of raw reads, Number of filtered reads, Number of aligned reads, Number of duplicate reads,
Median, mean and standard deviation of insert sizes of reads after alignment, percentage of bases
covered at X reads (%_bases_above_50 means the % of exons bases which have at least 50 reads)
whole genome or targeted percentage of bases covered at X reads (%_bases_above_50 means the % of exons
bases which have at least 50 reads). A TDF (.tdf) coverage track is also generated at this step
for easy visualization of coverage in the IGV browser.
"""
jobs = []
for sample in self.samples:
input_file_prefix = os.path.join("alignment", sample.name, sample.name + ".matefixed.sorted.")
input = input_file_prefix + "bam"
job = picard.collect_multiple_metrics(input, input_file_prefix + "all.metrics")
job.name = "picard_collect_multiple_metrics." + sample.name
jobs.append(job)
# Compute genome or target coverage with BVATools
job = bvatools.depth_of_coverage(
input,
input_file_prefix + "coverage.tsv",
bvatools.resolve_readset_coverage_bed(sample.readsets[0]),
other_options=config.param('bvatools_depth_of_coverage', 'other_options', required=False)
)
job.name = "bvatools_depth_of_coverage." + sample.name
jobs.append(job)
job = igvtools.compute_tdf(input, input + ".tdf")
job.name = "igvtools_compute_tdf." + sample.name
jobs.append(job)
return jobs
def get_metrics_jobs(self, readset):
jobs = []
input_file_prefix = readset.bam + '.'
input = input_file_prefix + "bam"
job = picard.collect_multiple_metrics(
input,
input_file_prefix + "metrics",
reference_sequence=readset.reference_file)
job.name = "picard_collect_multiple_metrics." + readset.name + ".met" + "." + readset.run + "." + readset.lane
jobs.append(job)
if readset.beds:
coverage_bed = readset.beds[0]
full_coverage_bed = (self.output_dir + os.sep + coverage_bed)
else:
coverage_bed = None
full_coverage_bed = None
if coverage_bed:
if (not os.path.exists(full_coverage_bed)) and \
(coverage_bed not in BwaRunProcessingAligner.downloaded_bed_files):
# Download the bed file
command = config.param('DEFAULT',
'fetch_bed_file_command').format(
output_directory=self.output_dir,
filename=coverage_bed)
job = Job([], [full_coverage_bed],
command=command,
name="bed_download." + coverage_bed)
BwaRunProcessingAligner.downloaded_bed_files.append(
coverage_bed)
jobs.append(job)
interval_list = re.sub("\.[^.]+$", ".interval_list", coverage_bed)
if interval_list not in BwaRunProcessingAligner.created_interval_lists:
# Create one job to generate the interval list from the bed file
ref_dict = os.path.splitext(
readset.reference_file)[0] + '.dict'
job = tools.bed2interval_list(ref_dict, full_coverage_bed,
interval_list)
job.name = "interval_list." + coverage_bed
BwaRunProcessingAligner.created_interval_lists.append(
interval_list)
jobs.append(job)
job = picard.calculate_hs_metrics(
input_file_prefix + "bam",
input_file_prefix + "metrics.onTarget.txt",
interval_list,
reference_sequence=readset.reference_file)
job.name = "picard_calculate_hs_metrics." + readset.name + ".hs" + "." + readset.run + "." + readset.lane
jobs.append(job)
jobs.extend(self.verify_bam_id(readset))
job = bvatools.depth_of_coverage(
input,
input_file_prefix + "metrics.targetCoverage.txt",
full_coverage_bed,
other_options=config.param('bvatools_depth_of_coverage',
'other_options',
required=False),
reference_genome=readset.reference_file)
job.name = "bvatools_depth_of_coverage." + readset.name + ".doc" + "." + readset.run + "." + readset.lane
jobs.append(job)
return jobs