def summary(items):
data = items[0]
cutoff = dd.get_coverage_depth_min(data)
work_dir = dd.get_work_dir(data)
out_dir = utils.safe_makedir(os.path.join(work_dir, "coverage"))
coverage_bed = dd.get_coverage_regions(data)
priority_bed = dd.get_priority_regions(data)
batch = _get_group_batch(items)
assert batch, "Did not find batch for samples: %s" % ",".join([dd.get_sample_name(x) for x in items])
out_file = os.path.join(out_dir, "%s-coverage.db" % batch)
if not utils.file_exists(out_file):
if coverage_bed:
mini_coverage = bed.minimize(coverage_bed).fn
if priority_bed:
mini_priority = bed.minimize(priority_bed).fn
if coverage_bed and priority_bed:
combined_bed = bed.concat([mini_coverage, mini_priority]).fn
elif coverage_bed:
combined_bed = mini_coverage
elif priority_bed:
combined_bed = mini_priority
else: # no coverage or priority file has been set
return items
clean_bed = bedutils.clean_file(combined_bed, data) if len(combined_bed) > 0 else combined_bed.fn
bed_file = _uniquify_bed_names(clean_bed, out_dir, data)
if bed_file and utils.file_exists(bed_file):
with file_transaction(data, out_file) as tx_out_file:
chanjo = os.path.join(os.path.dirname(sys.executable), "chanjo")
cmd = "{chanjo} --db {tx_out_file} build {bed_file}"
do.run(cmd.format(**locals()), "Prep chanjo database")
for data in items:
sample = dd.get_sample_name(data)
bam_file = data["work_bam"]
cmd = (
"{chanjo} annotate -s {sample} -g {batch} -c {cutoff} "
"{bam_file} {bed_file} | "
"{chanjo} --db {tx_out_file} import"
)
do.run(cmd.format(**locals()), "Chanjo coverage", data)
if bed_file:
os.remove(bed_file)
coverage = regions_coverage(out_file, batch, out_dir)
problem_regions = dd.get_problem_region_dir(data)
if problem_regions:
coverage = decorate_problem_regions(coverage, problem_regions)
out = []
for data in items:
if utils.file_exists(out_file):
data["coverage"] = {"summary": out_file, "all": coverage}
out.append([data])
return out
def junction2bed(junction_file):
"""
reformat the STAR junction file to BED3 format, one end of the splice junction per line
"""
base, _ = os.path.splitext(junction_file)
out_file = base + "-minimized.bed"
if file_exists(out_file):
return out_file
if not file_exists(junction_file):
return None
with file_transaction(out_file) as tx_out_file:
with open(junction_file) as in_handle:
with open(tx_out_file, "w") as out_handle:
for line in in_handle:
tokens = line.split()
chrom, sj1, sj2 = tokens[0:3]
if int(sj1) > int(sj2):
tmp = sj1
sj1 = sj2
sj2 = tmp
out_handle.write("\t".join([chrom, sj1, sj1]) + "\n")
out_handle.write("\t".join([chrom, sj2, sj2]) + "\n")
minimize = bed.minimize(tx_out_file)
minimize.saveas(tx_out_file)
return out_file
def junction2bed(junction_file):
"""
reformat the STAR junction file to BED3 format, one end of the splice junction per line
"""
base, _ = os.path.splitext(junction_file)
out_file = base + "-minimized.bed"
if file_exists(out_file):
return out_file
if not file_exists(junction_file):
return None
with file_transaction(out_file) as tx_out_file:
with open(junction_file) as in_handle:
with open(tx_out_file, "w") as out_handle:
for line in in_handle:
tokens = line.split()
chrom, sj1, sj2 = tokens[0:3]
if int(sj1) > int(sj2):
tmp = sj1
sj1 = sj2
sj2 = tmp
out_handle.write("\t".join([chrom, sj1, sj1]) + "\n")
out_handle.write("\t".join([chrom, sj2, sj2]) + "\n")
minimize = bed.minimize(tx_out_file)
minimize.saveas(tx_out_file)
return out_file
def junction2bed(junction_file):
"""
reformat the STAR junction file to BED3 format
"""
base, _ = os.path.splitext(junction_file)
out_file = base + "-minimized.bed"
if file_exists(out_file):
return out_file
minimized = bed.minimize(junction_file)
minimized.saveas(out_file)
return out_file
def summary(items):
data = items[0]
cutoff = dd.get_coverage_depth_min(data)
work_dir = dd.get_work_dir(data)
out_dir = utils.safe_makedir(os.path.join(work_dir, "coverage"))
coverage_bed = dd.get_coverage_regions(data)
priority_bed = dd.get_priority_regions(data)
batch = _get_group_batch(items)
assert batch, ("Did not find batch for samples: %s" %
",".join([dd.get_sample_name(x) for x in items]))
out_file = os.path.join(out_dir, "%s-coverage.db" % batch)
if not utils.file_exists(out_file):
mini_coverage = bed.minimize(coverage_bed).fn
mini_priority = bed.minimize(priority_bed).fn
combined_bed = bed.concat([mini_coverage, mini_priority]).fn
clean_bed = bedutils.clean_file(combined_bed, data) if len(combined_bed) > 0 else combined_bed.fn
bed_file = _uniquify_bed_names(clean_bed, out_dir, data)
if utils.file_exists(bed_file):
with file_transaction(data, out_file) as tx_out_file:
chanjo = os.path.join(os.path.dirname(sys.executable), "chanjo")
cmd = ("{chanjo} --db {tx_out_file} build {bed_file}")
do.run(cmd.format(**locals()), "Prep chanjo database")
for data in items:
sample = dd.get_sample_name(data)
bam_file = data["work_bam"]
cmd = ("{chanjo} annotate -s {sample} -g {batch} -c {cutoff} "
"{bam_file} {bed_file} | "
"{chanjo} --db {tx_out_file} import")
do.run(cmd.format(**locals()), "Chanjo coverage", data)
os.remove(bed_file)
coverage = regions_coverage(out_file, batch, out_dir)
problem_regions = dd.get_problem_region_dir(data)
if problem_regions:
coverage = decorate_problem_regions(coverage, problem_regions)
out = []
for data in items:
if utils.file_exists(out_file):
data["coverage"] = {"summary": out_file,
"all": coverage}
out.append([data])
return out