def fix_fusion(ref_f, genome_fa, out_file, no_fix, secondary_flag=0,
denovo_flag=0):
"""
Realign fusion juncrions
"""
print('Start to fix fusion junctions...')
fa = check_fasta(genome_fa)
ref = parse_ref(ref_f, 2)
annotated_fusion_f = 'annotated_fusion.txt.tmp'
fusions, fusion_names, fixed_flag = fix_bed(annotated_fusion_f, ref, fa,
no_fix, denovo_flag)
total = 0
annotations = set()
fixed_fusion_f = out_file
if secondary_flag:
secondary_f = open('low_conf_%s' % out_file, 'w')
with open(fixed_fusion_f, 'w') as outf:
for fus in fusion_names:
reads = str(fusions[fus])
fixed = fixed_flag[fus]
if fixed > 0:
total += 1
fixed = str(fixed)
name = 'circular_RNA/' + reads
if fus.startswith('secondary'):
_, loc, strand, left_info, right_info = fus.split('|')
secondary_f.write('\t'.join([loc, name, fixed, strand,
left_info, right_info]))
secondary_f.write('\n')
continue
gene, iso, chrom, strand, index = fus.split()
starts, ends = ref['\t'.join([gene, iso, chrom, strand])]
exon_num = len(starts)
intron_num = exon_num - 1
if ',' in index: # back spliced exons
s, e = [int(x) for x in index.split(',')]
if strand == '+':
index_info = ','.join(str(x + 1) for x in range(s, e + 1))
else:
index_info = ','.join(str(exon_num - x)
for x in range(s, e + 1))
start = str(starts[s])
end = str(ends[e])
length = str(e - s + 1)
sizes, offsets = generate_bed(int(start), starts[s:(e + 1)],
ends[s:(e + 1)])
annotation_info = '\t'.join([chrom, start, end, sizes,
offsets])
# remove circular RNA info duplications in denovo mode
if denovo_flag and annotation_info in annotations:
continue
if s == 0:
left_intron = 'None'
else:
left_intron = '%s:%d-%d' % (chrom, ends[s - 1], starts[s])
if e == len(ends) - 1:
right_intron = 'None'
else:
right_intron = '%s:%d-%d' % (chrom, ends[e], starts[e + 1])
intron = '|'.join([left_intron, right_intron])
bed = '\t'.join([chrom, start, end, name, fixed, strand, start,
start, '0,0,0', length, sizes, offsets,
reads, 'circRNA', gene, iso, index_info,
intron])
else: # ciRNAs
index, start, end = index.split('|')
size = str(int(end) - int(start))
annotation_info = '\t'.join([chrom, start, end, size, '0'])
# remove circular RNA info duplications in denovo mode
if denovo_flag and annotation_info in annotations:
continue
index = int(index)
if strand == '+':
index_info = str(index + 1)
else:
index_info = str(intron_num - index)
intron = '%s:%d-%d' % (chrom, ends[index], starts[index + 1])
bed = '\t'.join([chrom, start, end, name, fixed, strand, start,
start, '0,0,0', '1', size, '0',
reads, 'ciRNA', gene, iso, index_info,
intron])
if denovo_flag: # in denovo mode
annotations.add(annotation_info)
outf.write(bed + '\n')
if secondary_flag:
secondary_f.close()
os.remove('annotated_fusion.txt.tmp')
print('Fixed %d fusion junctions!' % total)
def fix_fusion(ref_f, genome_fa, out_dir, no_fix, denovo_flag=0):
"""
Realign fusion juncrions
"""
print('Start to fix fusion junctions...')
fa = check_fasta(genome_fa)
ref = parse_ref(ref_f, 2)
annotated_fusion_f = '%s/annotated_fusion.txt' % out_dir
fusions, fusion_names, fixed_flag = fix_bed(annotated_fusion_f, ref, fa,
no_fix, denovo_flag)
total = 0
annotations = set()
fixed_fusion_f = '%s/circ_fusion.txt' % out_dir
with open(fixed_fusion_f, 'w') as outf:
for fus in fusion_names:
reads = str(fusions[fus])
fixed = fixed_flag[fus]
if fixed > 0:
total += 1
fixed = str(fixed)
name = 'circular_RNA/' + reads
gene, iso, chrom, strand, index = fus.split()
starts, ends = ref['\t'.join([gene, iso, chrom, strand])]
exon_num = len(starts)
intron_num = exon_num - 1
if ',' in index: # back spliced exons
s, e = [int(x) for x in index.split(',')]
if strand == '+':
index_info = ','.join(str(x + 1) for x in xrange(s, e + 1))
else:
index_info = ','.join(str(exon_num - x)
for x in xrange(s, e + 1))
start = str(starts[s])
end = str(ends[e])
length = str(e - s + 1)
sizes, offsets = generate_bed(int(start), starts[s:(e + 1)],
ends[s:(e + 1)])
annotation_info = '\t'.join([chrom, start, end, sizes,
offsets])
# remove circular RNA info duplications in denovo mode
if denovo_flag and annotation_info in annotations:
continue
if s == 0:
left_intron = 'None'
else:
left_intron = '%s:%d-%d' % (chrom, ends[s - 1], starts[s])
if e == len(ends) - 1:
right_intron = 'None'
else:
right_intron = '%s:%d-%d' % (chrom, ends[e], starts[e + 1])
intron = '|'.join([left_intron, right_intron])
bed = '\t'.join([chrom, start, end, name, fixed, strand, start,
start, '0,0,0', length, sizes, offsets,
reads, 'circRNA', gene, iso, index_info,
intron])
else: # ciRNAs
index, start, end = index.split('|')
size = str(int(end) - int(start))
annotation_info = '\t'.join([chrom, start, end, size, '0'])
# remove circular RNA info duplications in denovo mode
if denovo_flag and annotation_info in annotations:
continue
index = int(index)
if strand == '+':
index_info = str(index + 1)
else:
index_info = str(intron_num - index)
intron = '%s:%d-%d' % (chrom, ends[index], starts[index + 1])
bed = '\t'.join([chrom, start, end, name, fixed, strand, start,
start, '0,0,0', '1', size, '0',
reads, 'ciRNA', gene, iso, index_info,
intron])
if denovo_flag: # in denovo mode
annotations.add(annotation_info)
outf.write(bed + '\n')
print('Fixed %d fusion junctions!' % total)