def sam_to_bam(input_fastq_files,
input_sam_file,
output_bam_file,
quals,
multihits,
pe_sr_mode=False,
softclip=True,
keep_unmapped=True):
samfh = pysam.Samfile(input_sam_file, "r")
num_unmapped = 0
num_multihits = 0
num_frags = 0
bamfh = pysam.Samfile(output_bam_file, "wb", template=samfh)
# setup fastq parsing
if softclip and (quals != SANGER_FORMAT):
kwargs = {"convert_quals": True, "qual_format": quals}
else:
kwargs = {"convert_quals": False}
fqiters = [
parse_fastq_record(open(fq), **kwargs) for fq in input_fastq_files
]
# handle single-read and paired-end
if len(fqiters) == 1:
reorder_func = fix_sr_alignment_ordering(samfh, fqiters[0])
else:
reorder_func = fix_alignment_ordering(samfh, fqiters, pe_sr_mode)
# iterate through buffer
for bufitems in reorder_func:
num_frags += 1
for bufitem in bufitems:
for r in bufitem.reads:
# softclip uses the fastq record to replace the sequence
# and quality scores of the read
if softclip:
soft_pad_read(bufitem.fqrec, r)
# keep statistics of unmapped/multimapped reads and
# suppress output if 'keep_unmapped' is False
if r.is_unmapped:
xm_tag = r.opt('XM')
if xm_tag < multihits:
num_unmapped += 1
if not keep_unmapped:
continue
else:
num_multihits += 1
bamfh.write(r)
for fqfh in fqiters:
fqfh.close()
bamfh.close()
samfh.close()
logging.debug("Found %d fragments" % (num_frags))
logging.debug("\t%d unmapped reads" % (num_unmapped))
logging.debug("\t%d multimapping (>%dX) reads" %
(num_multihits, multihits))
def sam_to_bam(input_fastq_files, input_sam_file, output_bam_file,
quals, multihits, pe_sr_mode=False, softclip=True,
keep_unmapped=True):
samfh = pysam.Samfile(input_sam_file, "r")
num_unmapped = 0
num_multihits = 0
num_frags = 0
bamfh = pysam.Samfile(output_bam_file, "wb", template=samfh)
# setup fastq parsing
if softclip and (quals != SANGER_FORMAT):
kwargs = {"convert_quals": True, "qual_format": quals}
else:
kwargs = {"convert_quals": False}
fqiters = [parse_fastq_record(open(fq), **kwargs) for fq in input_fastq_files]
# handle single-read and paired-end
if len(fqiters) == 1:
reorder_func = fix_sr_alignment_ordering(samfh, fqiters[0])
else:
reorder_func = fix_alignment_ordering(samfh, fqiters, pe_sr_mode)
# iterate through buffer
for bufitems in reorder_func:
num_frags += 1
for bufitem in bufitems:
for r in bufitem.reads:
# softclip uses the fastq record to replace the sequence
# and quality scores of the read
if softclip:
soft_pad_read(bufitem.fqrec, r)
# keep statistics of unmapped/multimapped reads and
# suppress output if 'keep_unmapped' is False
if r.is_unmapped:
xm_tag = r.opt('XM')
if xm_tag < multihits:
num_unmapped += 1
if not keep_unmapped:
continue
else:
num_multihits += 1
bamfh.write(r)
for fqfh in fqiters:
fqfh.close()
bamfh.close()
samfh.close()
logging.debug("Found %d fragments" % (num_frags))
logging.debug("\t%d unmapped reads" % (num_unmapped))
logging.debug("\t%d multimapping (>%dX) reads" %
(num_multihits, multihits))
def sam_stdin_to_bam(output_bam_file,
input_fastq_file,
multihits,
is_paired=True,
keep_unmapped=True):
samfh = pysam.Samfile("-", "r")
bamfh = pysam.Samfile(output_bam_file, "wb", template=samfh)
num_unmapped = 0
num_multihits = 0
if is_paired:
for pe_reads in fix_pe_alignment_ordering(samfh,
open(input_fastq_file),
is_paired=is_paired):
for reads in pe_reads:
for r in reads:
if r.is_unmapped:
xm_tag = r.opt('XM')
if xm_tag < multihits:
num_unmapped += 1
if not keep_unmapped:
continue
num_multihits += 1
bamfh.write(r)
else:
for reads in fix_sr_alignment_ordering(samfh, open(input_fastq_file)):
for r in reads:
if r.is_unmapped:
xm_tag = r.opt('XM')
if xm_tag < multihits:
num_unmapped += 1
if not keep_unmapped:
continue
num_multihits += 1
bamfh.write(r)
bamfh.close()
samfh.close()
logging.debug("[SAMTOBAM] Filtered %d unmapped reads" % (num_unmapped))
logging.debug("[SAMTOBAM] Found %d multimapping (>%d) reads" %
(num_multihits, multihits))
logging.info("[SAMTOBAM] Finished converting SAM -> BAM")
def sam_stdin_to_bam(output_bam_file, input_fastq_file, multihits,
is_paired=True, keep_unmapped=True):
samfh = pysam.Samfile("-", "r")
bamfh = pysam.Samfile(output_bam_file, "wb", template=samfh)
num_unmapped = 0
num_multihits = 0
if is_paired:
for pe_reads in fix_pe_alignment_ordering(samfh,
open(input_fastq_file),
is_paired=is_paired):
for reads in pe_reads:
for r in reads:
if r.is_unmapped:
xm_tag = r.opt('XM')
if xm_tag < multihits:
num_unmapped += 1
if not keep_unmapped:
continue
num_multihits += 1
bamfh.write(r)
else:
for reads in fix_sr_alignment_ordering(samfh, open(input_fastq_file)):
for r in reads:
if r.is_unmapped:
xm_tag = r.opt('XM')
if xm_tag < multihits:
num_unmapped += 1
if not keep_unmapped:
continue
num_multihits += 1
bamfh.write(r)
bamfh.close()
samfh.close()
logging.debug("[SAMTOBAM] Filtered %d unmapped reads" % (num_unmapped))
logging.debug("[SAMTOBAM] Found %d multimapping (>%d) reads" %
(num_multihits, multihits))
logging.info("[SAMTOBAM] Finished converting SAM -> BAM")
