def bed_tofasta(bed,
ref_fasta,
min_size=50,
stranded=True,
include_name=False,
out=sys.stdout):
if not os.path.exists('%s.fai' % ref_fasta):
pysam.faidx(ref_fasta)
fasta = pysam.Fastafile(ref_fasta)
refs = set()
with open('%s.fai' % ref_fasta) as f:
for line in f:
refs.add(line.split('\t')[0].strip())
name = ''
for region in bed:
if include_name:
name = '%s|' % (region.name.strip())
if region.end - region.start >= min_size and region.chrom in refs:
seq = fasta.fetch(region.chrom, region.start, region.end)
if stranded and region.strand:
if region.strand == '-':
seq = revcomp(seq)
out.write('>%s%s:%d-%d[%s]\n%s\n' %
(name, region.chrom, region.start, region.end,
region.strand, seq))
else:
out.write('>%s%s:%d-%d\n%s\n' %
(name, region.chrom, region.start, region.end, seq))
fasta.close()
def bed_tofasta(bed, ref_fasta, min_size=50, stranded=True, include_name=False, out=sys.stdout):
if not os.path.exists('%s.fai' % ref_fasta):
pysam.faidx(ref_fasta)
fasta = pysam.Fastafile(ref_fasta)
refs = set()
with open('%s.fai' % ref_fasta) as f:
for line in f:
refs.add(line.split('\t')[0].strip())
name = ''
for region in bed:
if include_name:
name = '%s|' % (region.name.strip())
if region.end - region.start >= min_size and region.chrom in refs:
seq = fasta.fetch(region.chrom, region.start, region.end)
if stranded and region.strand:
if region.strand == '-':
seq = revcomp(seq)
out.write('>%s%s:%d-%d[%s]\n%s\n' % (name, region.chrom, region.start, region.end, region.strand, seq))
else:
out.write('>%s%s:%d-%d%s\n%s\n' % (name, region.chrom, region.start, region.end, seq))
fasta.close()
def alleles(self):
alts = []
for alt in self.observed.split('/'):
if alt != '-' and self.strand == '-':
alt = revcomp(alt)
alts.append(alt)
return alts
def alleles(self):
alts = []
for alt in self.observed.split('/'):
if alt != '-' and self.strand == '-':
alt = revcomp(alt)
alts.append(alt)
return alts
def alleles(self):
alts = []
for alt in self.observed.split("/"):
if alt != "-" and self.strand == "-":
alt = revcomp(alt)
alts.append(alt)
return alts
def write_fasta(read, out=sys.stdout, colorspace=False):
if colorspace:
seq = read.opt('CS')
else:
seq = read.seq
if not read.is_unmapped and read.is_reverse:
if colorspace:
seq = seq[::-1]
else:
seq = revcomp(seq)
out.write('>%s\n%s\n' % (read.qname, seq))
def write_fasta(read, out=sys.stdout, colorspace=False):
if colorspace:
seq = read.opt('CS')
else:
seq = read.seq
if not read.is_unmapped and read.is_reverse:
if colorspace:
seq = seq[::-1]
else:
seq = revcomp(seq)
out.write('>%s\n%s\n' % (read.qname, seq))
def write_fastq(read, out=sys.stdout, colorspace=False):
if colorspace:
seq = read.opt('CS')
qual = read.opt('CQ')
else:
seq = read.seq
qual = read.qual
if not read.is_unmapped and read.is_reverse:
if colorspace:
seq = seq[::-1]
else:
seq = revcomp(seq)
qual = qual[::-1]
out.write('@%s\n%s\n+\n%s\n' % (read.qname, seq, qual))
def write_fastq(read, out=sys.stdout, colorspace=False):
if colorspace:
seq = read.opt('CS')
qual = read.opt('CQ')
else:
seq = read.seq
qual = read.qual
if not read.is_unmapped and read.is_reverse:
if colorspace:
seq = seq[::-1]
else:
seq = revcomp(seq)
qual = qual[::-1]
out.write('@%s\n%s\n+\n%s\n' % (read.qname, seq, qual))
def check_tags(barcodes, seq, edit, pos, allow_revcomp=False, verbose=False):
"""
For each barcode, pull out the appropriate 5' or 3' sub sequence from {seq}. Then
run a local alignment of the barcode to the subseq. If a good match is found, return
it; otherwise, find the best match and return that.
returns a multi-tuple:
valid, (tag, valid_seq, edits, reason_if_fail)
For the alignments, the reference is the barcode, the query is the subset of the read
that is possibly the barcode (5'/3' subseq)
"""
best = None
# check perfect matches first...
# for tag in barcodes:
# barcodeseq, orientation = barcodes[tag]
# if orientation == '5':
# if seq[:len(barcodeseq)] == barcodeseq:
# return True, (tag, seq[len(barcodeseq):], 0, '')
# else:
# if seq[-len(barcodeseq):] == barcodeseq:
# return True, (tag, seq[:-len(barcodeseq)], 0, '')
for tag in barcodes:
barcodeseq, orientation, strip = barcodes[tag]
if orientation == "5":
testseq = seq[: len(barcodeseq) + edit + pos]
else:
testseq = seq[-1 * (len(barcodeseq) + edit + pos) :]
aln = sw.align(barcodeseq, testseq)
valid, reason = _tag_aln_check(aln, len(testseq), len(barcodeseq), orientation, edit, pos)
if verbose:
print "Testing tag: %s vs %s" % (str(barcodes[tag]), testseq)
aln.dump()
print valid, reason
if valid:
return True, (tag, aln, True, "")
if not best or aln.score > best[1].score:
best = (tag, aln, True, reason)
if allow_revcomp:
if orientation == "5":
testseq = seq[-1 * (len(barcodeseq) + edit + pos) :]
else:
testseq = seq[: len(barcodeseq) + edit + pos]
aln = sw.align(revcomp(barcodeseq), testseq)
valid, reason = _tag_aln_check(
aln, len(testseq), len(barcodeseq), "5" if orientation == "3" else "3", edit, pos
)
if verbose:
print "Testing tag: %s [rc] vs %s" % (str(barcodes[tag]), testseq)
aln.dump()
print valid, reason
if valid:
return True, (tag, aln, False, "")
if not best or aln.score > best[1].score:
best = (tag, aln, False, reason)
if verbose:
print "BEST: ", best
best[1].dump()
return False, best
def check_tags(barcodes, seq, edit, pos, allow_revcomp=False, verbose=False):
'''
For each barcode, pull out the appropriate 5' or 3' sub sequence from {seq}. Then
run a local alignment of the barcode to the subseq. If a good match is found, return
it; otherwise, find the best match and return that.
returns a multi-tuple:
valid, (tag, valid_seq, edits, reason_if_fail)
For the alignments, the reference is the barcode, the query is the subset of the read
that is possibly the barcode (5'/3' subseq)
'''
best = None
# check perfect matches first...
# for tag in barcodes:
# barcodeseq, orientation = barcodes[tag]
# if orientation == '5':
# if seq[:len(barcodeseq)] == barcodeseq:
# return True, (tag, seq[len(barcodeseq):], 0, '')
# else:
# if seq[-len(barcodeseq):] == barcodeseq:
# return True, (tag, seq[:-len(barcodeseq)], 0, '')
for tag in barcodes:
barcodeseq, orientation, strip = barcodes[tag]
if orientation == '5':
testseq = seq[:len(barcodeseq) + edit + pos]
else:
testseq = seq[-1 * (len(barcodeseq) + edit + pos):]
aln = sw.align(barcodeseq, testseq)
valid, reason = _tag_aln_check(aln, len(testseq), len(barcodeseq),
orientation, edit, pos)
if verbose:
print 'Testing tag: %s vs %s' % (str(barcodes[tag]), testseq)
aln.dump()
print valid, reason
if valid:
return True, (tag, aln, True, '')
if not best or aln.score > best[1].score:
best = (tag, aln, True, reason)
if allow_revcomp:
if orientation == '5':
testseq = seq[-1 * (len(barcodeseq) + edit + pos):]
else:
testseq = seq[:len(barcodeseq) + edit + pos]
aln = sw.align(revcomp(barcodeseq), testseq)
valid, reason = _tag_aln_check(aln, len(testseq), len(barcodeseq),
'5' if orientation == '3' else '3',
edit, pos)
if verbose:
print 'Testing tag: %s [rc] vs %s' % (str(
barcodes[tag]), testseq)
aln.dump()
print valid, reason
if valid:
return True, (tag, aln, False, '')
if not best or aln.score > best[1].score:
best = (tag, aln, False, reason)
if verbose:
print 'BEST: ', best
best[1].dump()
return False, best