def ssw_check_parasail(query, target):
user_matrix = parasail.matrix_create("ACGT", 2, -2)
alignment_info = {}
for sQId, sQSeq, sQQual in read(query):
sQSeq_r = reverse_complement(sQSeq)
for sTId, sTSeq, STQual in read(target):
res = parasail.ssw(sQSeq, sTSeq, 3, 1, user_matrix)
res_r = parasail.ssw(sQSeq_r, sTSeq, 3, 1, user_matrix)
if res.score1 > res_r.score1:
score = res.score1
qstart, qend = res.read_begin1 + 1., res.read_end1 + 1
tstart, tend = res.ref_begin1 + 1, res.ref_end1 + 1
strand = '+'
else:
score = res_r.score1
qstart, qend = len(sQSeq) - res_r.read_end1, len(
sQSeq) - res_r.read_begin1
tstart, tend = res_r.ref_begin1 + 1, res_r.ref_end1 + 1
strand = '-'
alignment_info[sTId] = [
score,
int(qstart),
int(qend),
int(tstart),
int(tend), strand
]
return (alignment_info)
def get_refined_bp_sbnd(tconsensus, fasta_file_h, tchr, tstart, tend, tdir, hout_log, margin = 200):
tconsensus_part = tconsensus[:1000] if len(tconsensus) > 1000 else tconsensus
ref_len = fasta_file_h.get_reference_length(tchr)
if tstart < 1: tstart = 1
if ref_len < tend: tend = ref_len
if tdir == '+':
qseq = fasta_file_h.fetch(tchr, max(int(tstart) - margin, 0), int(tend))
else:
qseq = fasta_file_h.fetch(tchr, max(int(tstart) - 1, 0), int(tend) + margin)
qseq = reverse_complement(qseq)
user_matrix = parasail.matrix_create("ACGT", 1, -2)
res = parasail.ssw(qseq, tconsensus, 3, 1, user_matrix)
if res is None:
logger.debug(f"Alignment for breakpoint localization failed for {tchr},{tstart},{tend},{tdir}")
return None
if tdir == '+':
bp_pos_reference = tend - (len(qseq) - res.read_end1 - 1)
else:
bp_pos_reference = tstart + (len(qseq) - res.read_end1 - 1)
tconsensus_after = tconsensus[(res.ref_end1 + 1):]
return (bp_pos_reference, tconsensus_after)
def generate_paf_file(query_fasta, target_fasta, output_file):
user_matrix = parasail.matrix_create("ACGT", 2, -2)
with open(target_fasta, 'r') as hin:
for line in hin:
if line.startswith('>'):
tid = line.rstrip('\n').split(' ')[0].lstrip('>')
else:
tseq = line.rstrip('\n')
with open(query_fasta, 'r') as hin, open(output_file, 'w') as hout:
for line in hin:
if line.startswith('>'):
qid = line.rstrip('\n').lstrip('>')
else:
qseq = line.rstrip('\n')
res = parasail.ssw(qseq, tseq, 3, 1, user_matrix)
if res is not None:
print("%s\t%d\t%d\t%d\t+\t%s\t%d\t%d\t%d\t*\t*\t60" %
(qid, len(qseq), res.read_begin1, res.read_end1, tid,
len(tseq), res.ref_begin1, res.ref_end1),
file=hout)
else:
logger.warning(
f'Error occured in the alignment of {qid} and {tid} via parasail'
)
def generate_paf_file(self, query_fasta, target_fasta, output_file):
user_matrix = parasail.matrix_create("ACGT", 2, -2)
paf_rec_count = 0
with open(target_fasta, 'r') as hin:
for line in hin:
if line.startswith('>'):
tid = line.rstrip('\n').split(' ')[0].lstrip('>')
else:
tseq = line.rstrip('\n')
with open(query_fasta, 'r') as hin, open(output_file, 'w') as hout:
for line in hin:
if line.startswith('>'):
qid = line.rstrip('\n').lstrip('>')
else:
qseq = line.rstrip('\n')
res = parasail.ssw(qseq, tseq, 3, 1, user_matrix)
if res is not None:
print(f"{qid}\t{len(qseq)}\t{res.read_begin1}\t{res.read_end1}\t+\t" +
f"{tid}\t{len(tseq)}\t{res.ref_begin1}\t{res.ref_end1}\t*\t*\t60", file = hout)
paf_rec_count = paf_rec_count + 1
else:
self.parasail_error.append((qid, tid))
return(paf_rec_count)
def ssw_alignment(s1,
s2,
match_score=2,
mismatch_penalty=-2,
opening_penalty=3,
gap_ext=1):
user_matrix = parasail.matrix_create("ACGT", match_score, mismatch_penalty)
result = parasail.ssw(s1, s2, opening_penalty, gap_ext, user_matrix)
print(result, type(result), dir(result))
print(dir(result))
for attr, value in result.__dict__.items():
print(attr, value)
# print(result.ref_begin1, result.ref_end1, result.read_begin1, result.read_end1)
# print()
return s1_alignment, s2_alignment, cigar_string, cigar_tuples, result.score
def generate_paf_file(query_fasta, target_fasta, output_file):
user_matrix = parasail.matrix_create("ACGT", 2, -2)
with open(target_fasta, 'r') as hin:
for line in hin:
if line.startswith('>'):
tid = line.rstrip('\n').lstrip('>')
else:
tseq = line.rstrip('\n')
with open(query_fasta, 'r') as hin, open(output_file, 'w') as hout:
for line in hin:
if line.startswith('>'):
qid = line.rstrip('\n').lstrip('>')
else:
qseq = line.rstrip('\n')
res = parasail.ssw(qseq, tseq, 3, 1, user_matrix)
print("%s\t%d\t%d\t%d\t+\t%s\t%d\t%d\t%d\t*\t*\t60" %
(qid, len(qseq), res.read_begin1, res.read_end1, tid,
len(tseq), res.ref_begin1, res.ref_end1),
file=hout)
"""
parser = argparse.ArgumentParser( description='Align inserts to a series of target sequences')
parser.add_argument('--inserts', required=True)
parser.add_argument('--targets', required=True)
args = parser.parse_args()
# read in the control sequences
control_seqs = list()
with open(args.targets) as control_fh:
count = 0
name = ""
for line in control_fh:
count += 1
if count % 2 == 0:
control_seqs.append([name, line.strip()])
else:
name = line.strip()
scoring_matrix = parasail.matrix_create("ACGT", 5, -1)
print("Read\tReadLen\tReadStart\tReadEnd\tRef\tRefLen\tRefStart\tRefEnd\tScore\tMatches")
for read in pysam.FastxFile(args.inserts):
# align this read against all oligos
for (control_name, control_sequence) in control_seqs:
#result = parasail.sw_stats_table_scan_16(read.sequence, control_sequence, 5, 4, scoring_matrix)
if len(read.sequence) == 0:
continue
result = parasail.ssw(read.sequence, control_sequence, 5, 4, scoring_matrix)
result2 = parasail.sw_stats_table_striped_16(read.sequence, control_sequence, 5, 4, scoring_matrix)
print(read.name+"\t"+str(len(read.sequence))+"\t"+str(result.read_begin1)+"\t"+str(result.read_end1)+"\t"+control_name+"\t"+str(len(control_sequence))+"\t"+str(result.ref_begin1)+"\t"+str(result.ref_end1)+"\t"+str(result.score1)+"\t"+str(result2.matches))
def filter_sv_insertion_match(self,
sv,
ins,
filter_item="Duplicate_with_insertion"):
# only apply when the first sv is not insertion and the second sv is insertion type
if len(sv.inseq) >= 100 or len(ins.inseq) < 100: return
if not (sv.chr1 == ins.chr1 and abs(sv.pos1 - ins.pos1) <= self.bp_dist_margin) and \
not (sv.chr2 == ins.chr2 and abs(sv.pos2 - ins.pos2) <= self.bp_dist_margin):
return
ins_seg = self.reference_h.fetch(
ins.chr1,
max(ins.pos1 - self.validate_seg_len - self.bp_dist_margin - 1, 0),
ins.pos1 - 1)
ins_seg = ins_seg + ins.inseq
ins_seg = ins_seg + self.reference_h.fetch(
ins.chr1, ins.pos2 - 1,
ins.pos2 + self.validate_seg_len + self.bp_dist_margin - 1)
if sv.dir1 == '+':
tseq = self.reference_h.fetch(
sv.chr1, max(sv.pos1 - self.validate_seg_len - 1, 0),
sv.pos1 - 1)
else:
tseq = self.reference_h.fetch(sv.chr1, sv.pos1 - 1,
sv.pos1 + self.validate_seg_len - 1)
tseq = reverse_complement(tseq)
if sv.dir1 == '+':
sv_seg = tseq + sv.inseq
else:
sv_seg = tseq + reverse_complement(sv.inseq)
if sv.dir2 == '-':
tseq = self.reference_h.fetch(sv.chr2, sv.pos2 - 1,
sv.pos2 + self.validate_seg_len - 1)
else:
tseq = self.reference_h.fetch(
sv.chr2, max(sv.pos2 - self.validate_seg_len - 1, 0),
sv.pos2 - 1)
tseq = reverse_complement(tseq)
sv_seg = sv_seg + tseq
user_matrix = parasail.matrix_create("ACGT", 2, -2)
res = parasail.ssw(sv_seg, ins_seg, 3, 1, user_matrix)
res_r = parasail.ssw(reverse_complement(sv_seg), ins_seg, 3, 1,
user_matrix)
if res.score1 > res_r.score1:
match_ratio = float(
res.score1) / (2 * (res.ref_end1 - res.ref_begin1 + 1))
if match_ratio > 0.75:
if res.read_begin1 < 0.1 * len(
sv_seg) and res.read_end1 > 0.9 * len(sv_seg):
sv.filter.append(filter_item)
return
else:
match_ratio = float(
res_r.score1) / (2 * (res_r.ref_end1 - res_r.ref_begin1 + 1))
if match_ratio > 0.75:
if res_r.read_begin1 < 0.1 * len(
sv_seg) and res_r.read_end1 > 0.9 * len(sv_seg):
sv.filter.append(filter_item)
return
return
