def _filter_by_indels(options, chunk):
"""Filters a list of SNPs and Indels, such that no SNP is closer to
an indel than the value set in options.min_distance_to_indels, and
such that no two indels too close. If two or more indels are within
this distance, the indel with the highest QUAL score is retained. When
no unique highest QUAL score exists, an arbitrary indel is retained
among those indels with the highest QUAL score. SNPs are filtered
based on prefiltered Indels."""
indels = set([vcf for vcf in chunk if vcfwrap.is_indel(vcf)])
distance_between = options.min_distance_between_indels
indel_blacklist = _group_indels_near_position(indels, distance_between)
distance_to = options.min_distance_to_indels
snp_blacklist = _group_indels_near_position(indels, distance_to)
for vcf in chunk:
if vcfwrap.is_indel(vcf):
blacklisted = indel_blacklist.get(vcf.pos + 1, [vcf])
if vcf is not _select_best_indel(blacklisted):
_mark_as_filtered(vcf, "W:%i" % distance_between)
elif (vcf.alt != ".") and (vcf.pos in snp_blacklist):
# TODO: How to handle heterozygous SNPs near
_mark_as_filtered(vcf, "w:%i" % distance_to)
def _filter_by_indels(options, chunk):
"""Filters a list of SNPs and Indels, such that no SNP is closer to
an indel than the value set in options.min_distance_to_indels, and
such that no two indels too close. If two or more indels are within
this distance, the indel with the highest QUAL score is retained. When
no unique highest QUAL score exists, an arbitrary indel is retained
among those indels with the highest QUAL score. SNPs are filtered
based on prefiltered Indels."""
indels = set([vcf for vcf in chunk if vcfwrap.is_indel(vcf)])
distance_between = options.min_distance_between_indels
indel_blacklist = _group_indels_near_position(indels, distance_between)
distance_to = options.min_distance_to_indels
snp_blacklist = _group_indels_near_position(indels, distance_to)
for vcf in chunk:
if vcfwrap.is_indel(vcf):
blacklisted = indel_blacklist.get(vcf.pos + 1, [vcf])
if vcf is not _select_best_indel(blacklisted):
_mark_as_filtered(vcf, "W=%i" % distance_between)
elif (vcf.alt != ".") and (vcf.pos in snp_blacklist):
# TODO: How to handle heterozygous SNPs near
_mark_as_filtered(vcf, "w=%i" % distance_to)
def build_region(options, genotype, bed):
# Note that bed.end is a past-the-end coordinate
start = max(0, bed.start - options.padding)
indels = []
sequence = ["N"] * (bed.end - start)
for vcf in filter_vcfs(genotype, bed.contig, start, bed.end):
if vcfwrap.is_indel(vcf):
indels.append(vcf)
else:
add_snp(vcf, vcf.pos - start, sequence)
if not options.ignore_indels:
for vcf in indels:
add_indel(options, bed, vcf, sequence)
offset = bed.start - start
length = bed.end - bed.start
truncated = sequence[offset:offset + length]
# Discard insertions after the last position
truncated[-1] = truncated[-1][:1]
return "".join(truncated)
def build_region(options, genotype, bed):
# Note that bed.end is a past-the-end coordinate
start = max(0, bed.start - options.padding)
indels = []
sequence = ["N"] * (bed.end - start)
for vcf in filter_vcfs(genotype, bed.contig, start, bed.end):
if vcfwrap.is_indel(vcf):
indels.append(vcf)
else:
add_snp(vcf, vcf.pos - start, sequence)
if not options.ignore_indels:
for vcf in indels:
add_indel(options, bed, vcf, sequence)
offset = bed.start - start
length = bed.end - bed.start
truncated = sequence[offset:offset + length]
# Discard insertions after the last position
truncated[-1] = truncated[-1][:1]
return "".join(truncated)
