def toJSON(self):
d = {}
for key in [
'variant_types',
'so_annotations',
'ref_freqs',
'ref_acs',
'ref_hom_hemi',
'annotations',
'genes',
'exclude_genes',
]:
if getattr(self, key):
d[key] = getattr(self, key)
if getattr(self, 'locations'):
locations = []
for location1, location2 in self.locations:
chrom1, pos1 = genomeloc.get_chr_pos(location1)
chrom2, pos2 = genomeloc.get_chr_pos(location2)
if chrom1 != chrom2:
raise ValueError("locations have different chromosomes: %s:%s %s:%s" % (chrom1, pos1, chrom2, pos2))
locations.append("%s:%s-%s" % (chrom1, pos1, pos2))
d['locations'] = locations
return d
def toJSON(self):
d = {}
for key in [
'variant_types',
'so_annotations',
'ref_freqs',
'annotations',
'genes',
'exclude_genes',
]:
if getattr(self, key):
d[key] = getattr(self, key)
if getattr(self, 'locations'):
locations = []
for location1, location2 in self.locations:
chrom1, pos1 = genomeloc.get_chr_pos(location1)
chrom2, pos2 = genomeloc.get_chr_pos(location2)
if chrom1 != chrom2:
raise ValueError(
"locations have different chromosomes: %s:%s %s:%s" %
(chrom1, pos1, chrom2, pos2))
locations.append("%s:%s-%s" % (chrom1, pos1, pos2))
d['locations'] = locations
return d
def family_coverage_gene(request, family, gene_id):
project_id = family.project.project_id
gene = get_reference().get_gene(gene_id)
gene_structure = get_reference().get_gene_structure(gene_id)
individuals = family.get_individuals()
indiv_ids = [i.indiv_id for i in individuals]
num_individuals = len(indiv_ids)
coding_regions = []
for c in get_coding_regions_from_gene_structure(gene_id, gene_structure):
coding_region = {}
coding_region['start'] = genomeloc.get_chr_pos(c.xstart)[1]
coding_region['stop'] = genomeloc.get_chr_pos(c.xstop)[1]
coding_region['gene_id'] = c.gene_id
coding_region['size'] = c.xstop - c.xstart + 1
coding_regions.append(coding_region)
coverages = {}
for individual in individuals:
coverages[
individual.indiv_id] = get_coverage_store().get_coverage_for_gene(
str(individual.pk), gene['gene_id'])
whole_gene = Counter({'callable': 0, 'low_coverage': 0, 'poor_mapping': 0})
for coverage_spec in coverages.values():
whole_gene['callable'] += coverage_spec['gene_totals']['callable']
whole_gene['low_coverage'] += coverage_spec['gene_totals'][
'low_coverage']
whole_gene['poor_mapping'] += coverage_spec['gene_totals'][
'poor_mapping']
gene_coding_size = 0
for c in coding_regions:
gene_coding_size += c['stop'] - c['start'] + 1
totalsize = gene_coding_size * num_individuals
whole_gene['ratio_callable'] = whole_gene['callable'] / float(totalsize)
whole_gene['ratio_low_coverage'] = whole_gene['low_coverage'] / float(
totalsize)
whole_gene['ratio_poor_mapping'] = whole_gene['poor_mapping'] / float(
totalsize)
whole_gene['gene_coding_size'] = gene_coding_size
return render(
request, 'coverage/family_coverage_gene.html', {
'project': family.project,
'family': family,
'gene': gene,
'coverages_json': json.dumps(coverages),
'whole_gene_json': json.dumps(whole_gene),
'coding_regions_json': json.dumps(coding_regions),
'indiv_ids_json': json.dumps(indiv_ids),
'individuals': individuals,
'whole_gene': whole_gene,
})
def family_coverage_gene(request, family, gene_id):
project_id = family.project.project_id
gene = get_reference().get_gene(gene_id)
gene_structure = get_reference().get_gene_structure(gene_id)
individuals = family.get_individuals()
indiv_ids = [i.indiv_id for i in individuals]
num_individuals = len(indiv_ids)
coding_regions = []
for c in get_coding_regions_from_gene_structure(gene_id, gene_structure):
coding_region = {}
coding_region['start'] = genomeloc.get_chr_pos(c.xstart)[1]
coding_region['stop'] = genomeloc.get_chr_pos(c.xstop)[1]
coding_region['gene_id'] = c.gene_id
coding_region['size'] = c.xstop-c.xstart+1
coding_regions.append(coding_region)
coverages = {}
for individual in individuals:
coverages[individual.indiv_id] = get_coverage_store().get_coverage_for_gene(
str(individual.pk),
gene['gene_id']
)
whole_gene = Counter({'callable': 0, 'low_coverage': 0, 'poor_mapping': 0})
for coverage_spec in coverages.values():
whole_gene['callable'] += coverage_spec['gene_totals']['callable']
whole_gene['low_coverage'] += coverage_spec['gene_totals']['low_coverage']
whole_gene['poor_mapping'] += coverage_spec['gene_totals']['poor_mapping']
gene_coding_size = 0
for c in coding_regions:
gene_coding_size += c['stop']-c['start']+1
totalsize = gene_coding_size*num_individuals
whole_gene['ratio_callable'] = whole_gene['callable'] / float(totalsize)
whole_gene['ratio_low_coverage'] = whole_gene['low_coverage'] / float(totalsize)
whole_gene['ratio_poor_mapping'] = whole_gene['poor_mapping'] / float(totalsize)
whole_gene['gene_coding_size'] = gene_coding_size
return render(request, 'coverage/family_coverage_gene.html', {
'project': family.project,
'family': family,
'gene': gene,
'coverages_json': json.dumps(coverages),
'whole_gene_json': json.dumps(whole_gene),
'coding_regions_json': json.dumps(coding_regions),
'indiv_ids_json': json.dumps(indiv_ids),
'individuals': individuals,
'whole_gene': whole_gene,
})
def get_multiple_variants(self,
project_id,
family_id,
xpos_ref_alt_tuples,
user=None):
"""
Get one or more specific variants in a family
Variant should be identifiable by xpos, ref, and alt
Note that ref and alt are just strings from the VCF (for now)
"""
variant_ids = []
for xpos, ref, alt in xpos_ref_alt_tuples:
chrom, pos = get_chr_pos(xpos)
variant_ids.append("%s-%s-%s-%s" % (chrom, pos, ref, alt))
results = self.get_elasticsearch_variants(
project_id,
family_id=family_id,
variant_id_filter=variant_ids,
user=user)
# make sure all variants in xpos_ref_alt_tuples were retrieved and are in the same order.
# Return None for tuples that weren't found in ES.
results_by_xpos_ref_alt = {}
for r in results:
results_by_xpos_ref_alt[(r.xpos, r.ref, r.alt)] = r
# create a list that's the same length as the input list of xpos_ref_alt_tuples, putting None for
# xpos-ref-alt's that weren't found in the elasticsearch index
results = [results_by_xpos_ref_alt.get(t) for t in xpos_ref_alt_tuples]
return results
def get_single_variant(self,
project_id,
family_id,
xpos,
ref,
alt,
user=None):
chrom, pos = get_chr_pos(xpos)
variant_id = "%s-%s-%s-%s" % (chrom, pos, ref, alt)
results = list(
self.get_elasticsearch_variants(project_id,
family_id=family_id,
variant_id_filter=[variant_id],
user=user,
include_all_consequences=True))
if not results:
return None
if len(results) > 1:
raise ValueError(
"Multiple variant records found for project: %s family: %s %s-%s-%s-%s: \n %s"
% (project_id, family_id, chrom, pos, ref, alt, "\n".join(
[pformat(v.toJSON()) for v in results])))
variant = results[0]
return variant
def get_multiple_variants(self, project_id, family_id, xpos_ref_alt_tuples, user=None):
"""
Get one or more specific variants in a family
Variant should be identifiable by xpos, ref, and alt
Note that ref and alt are just strings from the VCF (for now)
"""
variant_ids = []
for xpos, ref, alt in xpos_ref_alt_tuples:
chrom, pos = get_chr_pos(xpos)
if chrom == 'M':
chrom = 'MT'
variant_ids.append("%s-%s-%s-%s" % (chrom, pos, ref, alt))
results = self.get_elasticsearch_variants(project_id, family_id=family_id, variant_id_filter=variant_ids, user=user)
# make sure all variants in xpos_ref_alt_tuples were retrieved and are in the same order.
# Return None for tuples that weren't found in ES.
results_by_xpos_ref_alt = {}
for r in results:
results_by_xpos_ref_alt[(r.xpos, r.ref, r.alt)] = r
# create a list that's the same length as the input list of xpos_ref_alt_tuples, putting None for
# xpos-ref-alt's that weren't found in the elasticsearch index
results = [results_by_xpos_ref_alt.get(t) for t in xpos_ref_alt_tuples]
return results
def __init__(self, xpos, ref, alt):
self.xpos = xpos
self.ref = ref
self.alt = alt
# TODO: should be implemented in genomeloc.py
self.xposx = xpos
if len(ref) == 1 and len(alt) > 1: # insertion
self.xposx += len(alt) - 1
elif len(ref) > 1 and len(alt) == 1: # deletion
self.xposx -= 1
elif len(ref) > 1 and len(alt) > 1: # multi base sub
self.xposx += len(alt) - 1
chrom, pos = genomeloc.get_chr_pos(self.xpos)
self.chr = chrom
self.pos = pos
self.pos_end = self.xposx % 1e9
# TODO: feels like this should be an ordered dict
self.genotypes = {}
self.extras = {}
self.annotation = None
self.gene_ids = None
self.coding_gene_ids = None
self.vcf_id = None
self.vartype = 'snp' if len(ref) == 1 and len(alt) == 1 else 'indel'
def toJSON(self):
d = {}
for key in [
'variant_types', 'so_annotations', 'ref_freqs', 'annotations',
'genes', "exclude_genes"
]:
if getattr(self, key):
d[key] = getattr(self, key)
if getattr(self, 'locations'):
d['locations'] = [
"%s:%s-%s" % (genomeloc.get_chr_pos(locA)[0],
genomeloc.get_chr_pos(locA)[1],
genomeloc.get_chr_pos(locB)[1])
for locA, locB in self.locations
]
return d
def get_multiple_variants(self, project_id, family_id,
xpos_ref_alt_tuples):
"""
Get one or more specific variants in a family
Variant should be identifiable by xpos, ref, and alt
Note that ref and alt are just strings from the VCF (for now)
"""
variant_ids = []
for xpos, ref, alt in xpos_ref_alt_tuples:
chrom, pos = get_chr_pos(xpos)
variant_ids.append("%s-%s-%s-%s" % (chrom, pos, ref, alt))
cache_key = (project_id, family_id, tuple(xpos_ref_alt_tuples))
if cache_key in self._results_cache:
results = self._results_cache[cache_key]
else:
results = list(
self.get_elasticsearch_variants(project_id,
family_id=family_id,
variant_id_filter=variant_ids))
# make sure all variants in xpos_ref_alt_tuples were retrieved and are in the same order.
# Return None for tuples that weren't found in ES.
results_by_xpos_ref_alt = {}
for r in results:
results_by_xpos_ref_alt[(r.xpos, r.ref, r.alt)] = r
results = [
results_by_xpos_ref_alt.get(t) for t in xpos_ref_alt_tuples
]
self._results_cache[cache_key] = results
return results
def get_single_variant(self, project_id, family_id, xpos, ref, alt):
chrom, pos = get_chr_pos(xpos)
variant_id = "%s-%s-%s-%s" % (chrom, pos, ref, alt)
cache_key = (project_id, family_id, xpos, ref, alt)
if cache_key in self._results_cache:
results = self._results_cache[cache_key]
else:
results = list(
self.get_elasticsearch_variants(project_id,
family_id=family_id,
variant_id_filter=[variant_id
]))
self._results_cache[cache_key] = results
if not results:
return None
if len(results) > 1:
raise ValueError(
"Multiple variant records found for project: %s family: %s %s-%s-%s-%s: \n %s"
% (project_id, family_id, chrom, pos, ref, alt, "\n".join(
[pformat(v.toJSON()) for v in results])))
variant = results[0]
return variant
def toDict(self):
genes = [{ 'gene' : bg.gene_symbol, 'cds_dist': bg.cds_dist } for bg in self.breakpointgene_set.all()]
chr,pos = genomeloc.get_chr_pos(self.xpos)
return {
'xpos' : self.xpos,
'chr' : chr,
'pos' : pos,
'obs' : self.obs,
'sample_count' : self.sample_count,
'consensus' : self.consensus,
'indiv_id' : self.individual.indiv_id,
'genes' : genes,
}
def toList(self):
genes = [{ 'gene' : bg.gene_symbol, 'cds_dist': bg.cds_dist } for bg in self.breakpointgene_set.all()]
chr,pos = genomeloc.get_chr_pos(self.xpos)
return [
self.xpos,
chr,
pos,
self.obs,
self.sample_count,
self.consensus,
self.partner,
self.individual.indiv_id,
genes,
]
def toDict(self):
genes = [{
'gene': bg.gene_symbol,
'cds_dist': bg.cds_dist
} for bg in self.breakpointgene_set.all()]
chr, pos = genomeloc.get_chr_pos(self.xpos)
return {
'xpos': self.xpos,
'chr': chr,
'pos': pos,
'obs': self.obs,
'sample_count': self.sample_count,
'consensus': self.consensus,
'indiv_id': self.individual.indiv_id,
'genes': genes,
}
def get_single_variant(self, project_id, family_id, xpos, ref, alt, user=None):
chrom, pos = get_chr_pos(xpos)
if chrom == 'M':
chrom = 'MT'
variant_id = "%s-%s-%s-%s" % (chrom, pos, ref, alt)
results = list(self.get_elasticsearch_variants(project_id, family_id=family_id, variant_id_filter=[variant_id], user=user, include_all_consequences=True))
if not results:
return None
if len(results) > 1:
raise ValueError("Multiple variant records found for project: %s family: %s %s-%s-%s-%s: \n %s" % (
project_id, family_id, chrom, pos, ref, alt, "\n".join([pformat(v.toJSON()) for v in results])))
variant = results[0]
return variant
def toList(self):
genes = [{
'gene': bg.gene_symbol,
'cds_dist': bg.cds_dist
} for bg in self.breakpointgene_set.all()]
chr, pos = genomeloc.get_chr_pos(self.xpos)
return [
self.xpos,
chr,
pos,
self.obs,
self.sample_count,
self.consensus,
self.partner,
self.individual.indiv_id,
genes,
]
