def read_pair_info_iter(bam_in, target_regions, references):
while (True):
try:
keep_going = True
while (keep_going):
r1 = bam_in.next()
if not (r1.is_secondary):
keep_going = False
keep_going = True
while (keep_going):
r2 = bam_in.next()
if not (r2.is_secondary):
keep_going = False
except StopIteration:
break
name1 = r1.qname.split()[0]
name2 = r2.qname.split()[0]
assert (name1 == name2)
if r1.is_read1:
read1 = r1
read2 = r2
else:
read1 = r2
read2 = r1
if r1.tid == r2.tid and r2.pos < r1.pos:
pos2 = r1
pos1 = r2
else:
pos1 = r1
pos2 = r2
read_pair_info = {}
read_pair_info['name'] = name1
if read1.is_read1:
read1_index = 1
read2_index = 2
else:
read1_index = 2
read2_index = 1
read_pair_info['r1_index'] = read1_index
read_pair_info['r2_index'] = read2_index
read_pair_info[
'is_duplicate'] = read1.is_duplicate or read2.is_duplicate
read_pair_info['r1_mapped'] = not (read1.is_unmapped)
read_pair_info['r2_mapped'] = not (read2.is_unmapped)
read_pair_info['r1_mapq'] = read1.mapq
read_pair_info['r2_mapq'] = read2.mapq
if read1.tid == -1:
ref1 = ''
else:
ref1 = references[read1.tid]
read_pair_info['r1_chrom'] = ref1
read_pair_info['r1_pos'] = read1.pos
read_pair_info['r1_direct'] = not (read1.is_reverse)
read_pair_info['r1_min_qual'] = tk_fasta.get_min_qual(read1.qual)
read_pair_info['r1_seq_len'] = len(
read1.seq) if read1.seq is not None else 0
if read1.seq is not None and 'N' in read1.seq:
read_pair_info['r1_contains_N'] = True
else:
read_pair_info['r1_contains_N'] = False
if read2.tid == -1:
ref2 = ''
else:
ref2 = references[read2.tid]
read_pair_info['r2_chrom'] = ref2
read_pair_info['r2_pos'] = read2.pos
read_pair_info['r2_direct'] = not (read2.is_reverse)
read_pair_info['r2_min_qual'] = tk_fasta.get_min_qual(read2.qual)
read_pair_info['r2_seq_len'] = len(
read2.seq) if read2.seq is not None else 0
if read2.seq is not None and 'N' in read2.seq:
read_pair_info['r2_contains_N'] = True
else:
read_pair_info['r2_contains_N'] = False
#
# TODO: it would be preferable to add override ability in the tk_io calls,
# but jedna doesn't have its own tenkit branch
#
read_raw_bc = crdna_io.get_read_raw_barcode(read1)
if read_raw_bc is None:
read_raw_bc = ''
read_bc = crdna_io.get_read_barcode(read1)
if read_bc is None:
read_bc = ''
read_bc_qual = crdna_io.get_read_barcode_qual(read1)
read_sample_index = tk_io.get_read_sample_index(read1)
if read_sample_index is None:
read_sample_index = ''
read_sample_index_qual = tk_io.get_read_sample_index_qual(read1)
read_pair_info['10X_raw_bc_len'] = len(read_raw_bc)
read_pair_info['sample_index_len'] = len(read_sample_index)
read_pair_info['10X_raw_bc'] = read_raw_bc
read_pair_info['10X_called_bc'] = read_bc
read_pair_info['sample_index'] = read_sample_index
read_pair_info['10X_bc_min_qual'] = tk_fasta.get_min_qual(read_bc_qual)
read_pair_info['10X_bc_mean_qual'] = tk_fasta.get_mean_qual(
read_bc_qual)
if read_sample_index:
read_pair_info['sample_index_min_qual'] = tk_fasta.get_min_qual(
read_sample_index_qual)
read_pair_info['sample_index_mean_qual'] = tk_fasta.get_mean_qual(
read_sample_index_qual)
else:
read_pair_info['sample_index_min_qual'] = None
read_pair_info['sample_index_mean_qual'] = None
if not (ref1 == ref2) or target_regions is None:
read_pair_targ_dist = None
else:
left_pos = pos1.pos
right_pos = pos2.pos + (len(pos2.seq)
if pos2.seq is not None else 0)
regions = target_regions.get(ref1, None)
if regions is None:
# No targets on this chrom -- distance is large positive number
read_pair_targ_dist = 100000000
else:
read_pair_targ_dist = get_read_regions_dist(
left_pos, right_pos, regions)
read_pair_info['read_pair_targ_dist'] = read_pair_targ_dist
r1_mapped, r1_soft_clipped = cigar_count_bases(read1)
r2_mapped, r2_soft_clipped = cigar_count_bases(read2)
read_pair_info['mapped_bases'] = r1_mapped + r2_mapped
read_pair_info[
'soft_clipped_bases'] = r1_soft_clipped + r2_soft_clipped
if (not read1.is_unmapped
and not read2.is_unmapped) and read1.tid == read2.tid:
read_pair_info['insert_length'] = pos2.aend - pos1.pos
else:
read_pair_info['insert_length'] = None
read_pair_info['r1_q20_bases'] = tk_fasta.get_bases_qual(
read1.qual, 20)
read_pair_info['r1_q30_bases'] = tk_fasta.get_bases_qual(
read1.qual, 30)
read_pair_info['r2_q20_bases'] = tk_fasta.get_bases_qual(
read2.qual, 20)
read_pair_info['r2_q30_bases'] = tk_fasta.get_bases_qual(
read2.qual, 30)
read_pair_info['si_q20_bases'] = tk_fasta.get_bases_qual(
read_sample_index_qual, 20)
read_pair_info['si_q30_bases'] = tk_fasta.get_bases_qual(
read_sample_index_qual, 30)
read_pair_info['bc_q20_bases'] = tk_fasta.get_bases_qual(
read_bc_qual, 20)
read_pair_info['bc_q30_bases'] = tk_fasta.get_bases_qual(
read_bc_qual, 30)
yield read_pair_info
def min_qual_below(qual, threshold):
""" Return true if the min qual is below the threshold
qual: a read's qual string """
return threshold is not None and tk_fasta.get_min_qual(qual) < threshold