def get_distributions_by_type(records, variant_types, field, bins):
num_bins = len(bins)
counts = {}
types_set = set(variant_types)
for type in variant_types:
counts[type] = [0] * (num_bins + 1)
for record in records:
type = vu.get_sv_type(record, types_set)
val = vu.get_info_field(record, field)
idx = get_distribution_index(val, bins, num_bins)
counts[type][idx] += 1
return counts
def create_trees_from_records(records, variant_types, contigs, padding=0):
trees = {}
variant_types_set = set(variant_types)
for type in variant_types:
trees[type] = {}
for contig in contigs:
trees[type][contig] = IntervalTree()
for record in records:
type = vu.get_sv_type(record, variant_types_set)
contig = record.chrom
length = vu.get_record_length(record)
trees[type][contig].addi(record.start - padding,
record.start + length + padding)
return trees
def collect_evidence_fields(records, variant_types):
evidence_counts = {}
for variant_type in variant_types:
evidence_counts[variant_type] = {}
for evidence_type in EVIDENCE_TYPES:
evidence_counts[variant_type][evidence_type] = 0
variant_types_set = set(variant_types)
evidence_types_set = set(EVIDENCE_TYPES)
for record in records:
variant_type = vu.get_sv_type(record, variant_types_set)
evidence_types = vu.get_evidence_types(record, evidence_types_set)
for evidence_type in evidence_types:
evidence_counts[variant_type][evidence_type] += 1
return evidence_counts
def get_allele_frequency_counts(records, header, variant_types):
num_samples = float(len(header.samples))
allele_freq = {}
num_singletons = {}
types_set = set(variant_types)
# Don't calculate MCNV AF since non-ref alleles cannot be determined without chromosome ploidy
af_types = types_set - set(["CNV"])
for type in af_types:
allele_freq[type] = []
num_singletons[type] = 0
for record in records:
type = vu.get_sv_type(record, types_set)
if type not in af_types:
continue
af = 0
for sample in record.samples.values():
for val in sample["GT"]:
if val is not None and val > 0:
af += 1
if af == 1:
num_singletons[type] += 1
allele_freq[type].append(af / num_samples)
allele_freq_counts = {}
num_bins = len(AF_BINS)
for type in af_types:
allele_freq_counts[type] = [0] * (num_bins + 1)
for val in allele_freq[type]:
idx = get_distribution_index(val, AF_BINS, num_bins)
allele_freq_counts[type][idx] += 1
return allele_freq_counts, num_singletons
def get_count_by_type(records, variant_types):
counts = {}
types_set = set(variant_types)
for type in variant_types:
counts[type] = 0
for record in records:
type = vu.get_sv_type(record, types_set)
counts[type] += 1
return counts