def _merge_records(vcf, cpx_records, cpx_record_ids):
"""
r1, r2 : iter of pysam.VariantRecord
"""
def _next_record():
try:
return next(vcf)
except StopIteration:
return None
def _next_cpx():
try:
return cpx_records.popleft()
except IndexError:
return None
# Initialize merge
curr_record = _next_record()
curr_cpx = _next_cpx()
while curr_record is not None and curr_cpx is not None:
# Remove VCF records that were included in complex event
if curr_record.id in cpx_record_ids:
curr_record = _next_record()
continue
# Merge sort remaining
if curr_record.chrom == curr_cpx.chrom:
if curr_record.pos <= curr_cpx.pos:
yield curr_record
curr_record = _next_record()
else:
yield curr_cpx
curr_cpx = _next_cpx()
elif svu.is_smaller_chrom(curr_record.chrom, curr_cpx.chrom):
yield curr_record
curr_record = _next_record()
else:
yield curr_cpx
curr_cpx = _next_cpx()
# After one iterator is exhausted, return rest of other iterator
if curr_record is None:
for cpx in itertools.chain([curr_cpx], cpx_records):
yield cpx
elif curr_cpx is None:
for record in itertools.chain([curr_record], vcf):
if record.id not in cpx_record_ids:
yield record
def _merge_records(vcf, cpx_records, cpx_record_ids):
"""
r1, r2 : iter of pysam.VariantRecord
"""
def _next_record():
# Skip VCF records that were included in complex event
# get next record that's not already present in cpx_record_idss
_rec = next(vcf, None)
while _rec is not None and _rec.id in cpx_record_ids:
_rec = next(vcf, None)
return _rec
def _next_cpx():
try:
return cpx_records.popleft()
except IndexError:
return None
# Initialize merge
curr_record = _next_record()
curr_cpx = _next_cpx()
while curr_record is not None and curr_cpx is not None:
# Merge sort records not in complex event
if curr_record.chrom == curr_cpx.chrom:
if curr_record.pos <= curr_cpx.pos:
yield curr_record
curr_record = _next_record()
else:
yield curr_cpx
curr_cpx = _next_cpx()
elif svu.is_smaller_chrom(curr_record.chrom, curr_cpx.chrom):
yield curr_record
curr_record = _next_record()
else:
yield curr_cpx
curr_cpx = _next_cpx()
# At least one iterator is exhausted, return rest of other iterator, if any
while curr_record is not None:
yield curr_record
curr_record = _next_record()
while curr_cpx is not None:
yield curr_cpx
curr_cpx = _next_cpx()
def standardize_info(self, std_rec, raw_rec):
"""
Standardize Lumpy record.
1) Add CHR2, END
"""
std_rec.info['SVTYPE'] = raw_rec.info['SVTYPE']
# Strip per-strand counts
std_rec.info['STRANDS'] = raw_rec.info['STRANDS'][0].split(':')[0]
# Parse CHR2 and END
if std_rec.info['SVTYPE'] == 'BND':
chr2, end = parse_bnd_pos(std_rec.alts[0])
# swap chr2/chrom, pos/end, and reverse strandedness
if not is_smaller_chrom(std_rec.chrom, chr2):
std_rec.pos, end = end, std_rec.pos
std_rec.chrom, chr2 = chr2, std_rec.chrom
std_rec.info['STRANDS'] = std_rec.info['STRANDS'][::-1]
else:
chr2, end = raw_rec.chrom, raw_rec.stop
std_rec.info['CHR2'] = chr2
std_rec.stop = end
# Add SVLEN
if std_rec.chrom == std_rec.info['CHR2']:
std_rec.info['SVLEN'] = end - std_rec.pos
else:
std_rec.info['SVLEN'] = -1
std_rec.info['ALGORITHMS'] = ['smoove']
return std_rec
def standardize_info(self, std_rec, raw_rec):
"""
Standardize Delly record.
1) Rename 'TRA' to 'BND'.
2) Swap CHROM and CHR2 in translocations.
3) Add END.
4) Rename 'CT' to 'STRANDS' and convert notation.
5) Compute SVLEN.
6) Add ALGORITHMS.
7) Standardize ALT to VCF spec.
"""
# Rename TRA to BND
svtype = raw_rec.info['SVTYPE']
if svtype == 'TRA':
svtype = 'BND'
std_rec.info['SVTYPE'] = svtype
END = std_rec.stop
if "<" not in std_rec.alts[0]:
std_rec.alts = ('<' + svtype + '>', )
std_rec.ref = "N"
std_rec.stop = END
# Convert strandedness notation
raw_strands = raw_rec.info['CT']
if raw_strands == '5to3':
strands = '-+'
elif raw_strands == '3to3':
strands = '++'
elif raw_strands == '5to5':
strands = '--'
elif raw_strands == '3to5':
strands = '+-'
elif raw_strands == 'NtoN' and svtype == 'INS':
strands = '+-'
else:
msg = 'Improper strands ({0}) in record {1}'
raise Exception(msg.format(raw_strands, raw_rec.id))
std_rec.info['STRANDS'] = strands
pos, end = raw_rec.pos, raw_rec.stop
if pos == 0:
pos = 1
# Swap CHR2/CHROM if necessary and update ALT
if svtype == 'BND':
chrom, chr2 = raw_rec.chrom, raw_rec.info['CHR2']
# swap chr2/chrom, pos/end, and reverse strandedness
if not is_smaller_chrom(chrom, chr2):
if end == 0:
end = 1
std_rec.pos, end = end, pos
std_rec.chrom, chr2 = chr2, chrom
std_rec.info['STRANDS'] = strands[::-1]
else:
chr2 = raw_rec.chrom
# Add CHR2 and END
std_rec.stop = end
std_rec.info['CHR2'] = chr2
# Add SVLEN
if std_rec.chrom == std_rec.info['CHR2']:
std_rec.info['SVLEN'] = std_rec.stop - std_rec.pos
else:
std_rec.info['SVLEN'] = -1
std_rec.info['ALGORITHMS'] = ['delly']
return std_rec
def __lt__(self, other):
if self.tup[0] == other.tup[0]:
return int(self.tup[1]) < int(other.tup[1])
else:
return is_smaller_chrom(self.tup[0], other.tup[0])
def __lt__(self, other):
if self.record.chrom == other.record.chrom:
return self.record.pos < other.record.pos
else:
return svu.is_smaller_chrom(self.record.chrom, other.record.chrom)
def standardize_info(self, std_rec, raw_rec):
"""
Standardize Manta record.
1) Replace colons in ID with underscores (otherwise breaks VCF parsing)
2) Define CHR2 and END
3) Add strandedness
4) Add SVLEN
"""
# Colons in the ID can break parsing
std_rec.id = std_rec.id.replace(':', '_')
svtype = raw_rec.info['SVTYPE']
std_rec.info['SVTYPE'] = svtype
# Define CHR2 and END
if svtype == 'BND':
chr2, end = parse_bnd_pos(raw_rec.alts[0])
chrom, pos = raw_rec.chrom, raw_rec.pos
if not is_smaller_chrom(chrom, chr2):
pos, end = end, pos
chrom, chr2 = chr2, chrom
std_rec.pos = pos
std_rec.chrom = chrom
elif svtype == 'INS':
chr2 = raw_rec.chrom
end = raw_rec.pos + 1
else:
chr2 = raw_rec.chrom
end = raw_rec.stop
std_rec.info['CHR2'] = chr2
std_rec.stop = end
# Strand parsing
if svtype == 'INV':
if 'INV3' in raw_rec.info.keys():
strands = '++'
else:
strands = '--'
elif svtype == 'BND':
strands = parse_bnd_strands(raw_rec.alts[0])
elif svtype == 'DEL':
strands = '+-'
elif svtype == 'DUP':
strands = '-+'
elif svtype == 'INS':
strands = '+-'
if not is_smaller_chrom(std_rec.chrom, std_rec.info['CHR2']):
strands = strands[::-1]
std_rec.info['STRANDS'] = strands
if svtype == 'BND' and std_rec.chrom != std_rec.info['CHR2']:
std_rec.info['SVLEN'] = -1
elif svtype == 'INS':
std_rec.info['SVLEN'] = raw_rec.info.get('SVLEN', -1)
else:
std_rec.info['SVLEN'] = std_rec.stop - std_rec.pos
std_rec.info['ALGORITHMS'] = ['manta']
return std_rec
