def process_tophat_alignments(fastq_files,
bam_file,
gene_file,
max_fragment_length,
output_fastq_files,
output_bam_file,
unpaired=False,
suffix="/"):
# index genes
exon_intervals, exon_trees = build_exon_interval_trees(gene_file)
# open input files
bamfh = pysam.Samfile(bam_file, "rb")
if unpaired:
bam_iter = parse_unpaired_pe_reads(bamfh)
else:
bam_iter = parse_pe_reads(bamfh)
fastq_iters = [parse_fastq(open(fq)) for fq in fastq_files]
# open output files
outfq = [open(fq, "w") for fq in output_fastq_files]
outbamfh = pysam.Samfile(output_bam_file, "wb", template=bamfh)
# iterate through fastq files and bam file
try:
while True:
bam_pe_reads = bam_iter.next()
# synchronize fastq and bam and write unmapped reads to a file
is_unaligned = synchronize_bam_fastq(bam_pe_reads, fastq_iters,
outfq, suffix)
if is_unaligned:
continue
# if loop reaches this point then we have a paired-end
# read where both pairs align. now need to check if
# the alignment is discordant
tx_concordant, gene_concordant = \
is_concordant(bamfh, bam_pe_reads, exon_intervals,
exon_trees, max_fragment_length)
if not gene_concordant:
for r in bam_pe_reads[0]:
outbamfh.write(r)
for r in bam_pe_reads[1]:
outbamfh.write(r)
except StopIteration:
pass
# finish remaining fastq lines
try:
while True:
fqreads = [it.next() for it in fastq_iters]
print >> outfq[0], fastq_to_string(fqreads[0])
print >> outfq[1], fastq_to_string(fqreads[1])
except StopIteration:
pass
return config.JOB_SUCCESS
def process_tophat_alignments(fastq_files, bam_file, gene_file,
max_fragment_length,
output_fastq_files,
output_bam_file,
unpaired=False,
suffix="/"):
# index genes
exon_intervals, exon_trees = build_exon_interval_trees(gene_file)
# open input files
bamfh = pysam.Samfile(bam_file, "rb")
if unpaired:
bam_iter = parse_unpaired_pe_reads(bamfh)
else:
bam_iter = parse_pe_reads(bamfh)
fastq_iters = [parse_fastq(open(fq)) for fq in fastq_files]
# open output files
outfq = [open(fq, "w") for fq in output_fastq_files]
outbamfh = pysam.Samfile(output_bam_file, "wb", template=bamfh)
# iterate through fastq files and bam file
try:
while True:
bam_pe_reads = bam_iter.next()
# synchronize fastq and bam and write unmapped reads to a file
is_unaligned = synchronize_bam_fastq(bam_pe_reads, fastq_iters,
outfq, suffix)
if is_unaligned:
continue
# if loop reaches this point then we have a paired-end
# read where both pairs align. now need to check if
# the alignment is discordant
tx_concordant, gene_concordant = \
is_concordant(bamfh, bam_pe_reads, exon_intervals,
exon_trees, max_fragment_length)
if not gene_concordant:
for r in bam_pe_reads[0]:
outbamfh.write(r)
for r in bam_pe_reads[1]:
outbamfh.write(r)
except StopIteration:
pass
# finish remaining fastq lines
try:
while True:
fqreads = [it.next() for it in fastq_iters]
print >>outfq[0], fastq_to_string(fqreads[0])
print >>outfq[1], fastq_to_string(fqreads[1])
except StopIteration:
pass
return config.JOB_SUCCESS
def write_pe_fastq(fqreads, outfq, suffix):
print >> outfq[0], fastq_to_string(fqreads[0], suffix="%s1" % (suffix))
print >> outfq[1], fastq_to_string(fqreads[1], suffix="%s2" % (suffix))
def write_pe_fastq(fqreads, outfq, suffix):
print >>outfq[0], fastq_to_string(fqreads[0], suffix="%s1" % (suffix))
print >>outfq[1], fastq_to_string(fqreads[1], suffix="%s2" % (suffix))
