def _create_tables(self, config, temp):
# Opening pipe/symlink created in _setup()
out = sys.stderr
if not self._print_stats:
out = open(os.path.join(temp, "pipe_coverage_%i.stdout" % id(self)), "w")
with pysam.Samfile(self._pipes["input_file"]) as samfile:
timer = BAMTimer(samfile, out = out)
intervals, region_names = self._get_intervals(temp, samfile)
mapping = self._open_handles(temp, samfile, intervals)
for read in samfile:
if read.is_unmapped or read.is_duplicate:
continue
rg = dict(read.tags).get("RG")
for handle in mapping[rg]:
handle.write(read)
timer.increment(read = read)
timer.finalize()
if not self._print_stats:
out.close()
for handle in self._handle.itervalues():
handle.close()
for proclst in self._procs.itervalues():
for proc in proclst:
if proc.wait() != 0:
raise RuntimeError("Error while running process: %i" % proc.wait())
return region_names
def process_file(handle, args):
timer = BAMTimer(handle, step=1000000)
counts = {}
region_template = build_region_template(args, handle)
for region in BAMRegionsIter(handle, args.regions):
if region.name is None:
# Trailing unmapped reads
continue
name = region.name
if not args.regions and (handle.nreferences > args.max_contigs):
name = '<Genome>'
region_table = get_region_table(counts, name, region_template)
for (_, records) in region:
for record in records:
readgroup = args.get_readgroup_func(record)
readgroup_table = region_table[readgroup]
process_record(readgroup_table, record, record.flag, region)
timer.increment(read=record)
timer.finalize()
print_table(args, handle, counts)
return 0
def _create_tables(self, config, temp):
# Opening pipe/symlink created in _setup()
out = sys.stderr
if not self._print_stats:
out = open(
os.path.join(temp, "pipe_coverage_%i.stdout" % id(self)), "w")
with pysam.Samfile(self._pipes["input_file"]) as samfile:
timer = BAMTimer(samfile, out=out)
intervals, region_names = self._get_intervals(temp, samfile)
mapping = self._open_handles(temp, samfile, intervals)
for read in samfile:
if read.is_unmapped or read.is_duplicate:
continue
rg = dict(read.tags).get("RG")
for handle in mapping[rg]:
handle.write(read)
timer.increment(read=read)
timer.finalize()
if not self._print_stats:
out.close()
for handle in self._handle.itervalues():
handle.close()
for proclst in self._procs.itervalues():
for proc in proclst:
if proc.wait() != 0:
raise RuntimeError("Error while running process: %i" %
proc.wait())
return region_names
def process_file(handle, args):
timer = BAMTimer(handle, step=1000000)
counts = {}
region_template = build_region_template(args, handle)
for region in BAMRegionsIter(handle, args.regions):
if region.name is None:
# Trailing unmapped reads
continue
name = region.name
if not args.regions and (handle.nreferences > args.max_contigs):
name = '<Genome>'
region_table = get_region_table(counts, name, region_template)
for (_, records) in region:
for record in records:
readgroup = args.get_readgroup_func(record)
readgroup_table = region_table[readgroup]
process_record(readgroup_table, record, record.flag, region)
timer.increment(read=record)
timer.finalize()
print_table(args, handle, counts)
return 0
def process_file(handle, args):
timer = BAMTimer(handle, step=1000000)
last_tid = 0
totals = build_totals_dict(args, handle)
rg_to_smlbid, smlbid_to_smlb = build_rg_to_smlbid_keys(args, handle)
template = [0] * len(smlbid_to_smlb)
for region in BAMRegionsIter(handle, args.regions):
if region.name is None:
# Trailing unmapped reads
continue
elif not args.regions and (handle.nreferences > args.max_contigs):
region.name = '<Genome>'
last_pos = 0
counts = collections.deque()
mapping = MappingToTotals(totals, region, smlbid_to_smlb)
for (position, records) in region:
mapping.process_counts(counts, last_pos, position)
for record in records:
timer.increment(read=record)
count_bases(args, counts, record, rg_to_smlbid, template)
if (region.tid, position) < (last_tid, last_pos):
sys.stderr.write("ERROR: Input BAM file is unsorted\n")
return 1
last_pos = position
last_tid = region.tid
# Process columns in region after last read
mapping.process_counts(counts, last_pos, float("inf"))
mapping.finalize()
timer.finalize()
print_table(handle, args, totals)
return 0
def process_file(handle, args):
timer = BAMTimer(handle, step=1000000)
last_tid = 0
totals = build_totals_dict(args, handle)
rg_to_smlbid, smlbid_to_smlb = build_rg_to_smlbid_keys(args, handle)
template = [0] * len(smlbid_to_smlb)
for region in BAMRegionsIter(handle, args.regions):
if region.name is None:
# Trailing unmapped reads
continue
elif not args.regions and (handle.nreferences > args.max_contigs):
region.name = '<Genome>'
last_pos = 0
counts = collections.deque()
mapping = MappingToTotals(totals, region, smlbid_to_smlb)
for (position, records) in region:
mapping.process_counts(counts, last_pos, position)
for record in records:
timer.increment(read=record)
count_bases(args, counts, record, rg_to_smlbid, template)
if (region.tid, position) < (last_tid, last_pos):
sys.stderr.write("ERROR: Input BAM file is unsorted\n")
return 1
last_pos = position
last_tid = region.tid
# Process columns in region after last read
mapping.process_counts(counts, last_pos, float("inf"))
mapping.finalize()
timer.finalize()
print_table(handle, args, totals)
return 0