def add_hgvs_allele(self, hgvs_allele):
"""parse and add the hgvs_allele to the bundle"""
hp = _get_hgvs_parser()
sv = hp.parse_hgvs_variant(hgvs_allele)
sequence_id = get_vmc_sequence_identifier(sv.ac)
self.identifiers[sequence_id].add(sv.ac)
if isinstance(sv.posedit.pos, hgvs.location.BaseOffsetInterval):
if sv.posedit.pos.start.is_intronic or sv.posedit.pos.end.is_intronic:
raise ValueError("Intronic HGVS variants are not supported".format(sv.posedit.edit.type))
if sv.posedit.edit.type == 'ins':
interval = models.Interval(start=sv.posedit.pos.start.base, end=sv.posedit.pos.start.base)
state = sv.posedit.edit.alt
elif sv.posedit.edit.type in ('sub', 'del', 'delins', 'identity'):
interval = models.Interval(start=sv.posedit.pos.start.base - 1, end=sv.posedit.pos.end.base)
if sv.posedit.edit.type == 'identity':
state = get_reference_sequence(sv.ac, sv.posedit.pos.start.base - 1, sv.posedit.pos.end.base)
else:
state = sv.posedit.edit.alt or ''
else:
raise ValueError("HGVS variant type {} is unsupported".format(sv.posedit.edit.type))
location = models.Location(sequence_id=sequence_id, interval=interval)
location.id = self._id_function(location)
self.locations[location.id] = location
allele = models.Allele(location_id=location.id, state=state)
allele.id = self._id_function(allele)
self.alleles[allele.id] = allele
return allele
def from_hgvs(hgvs_string):
hp = _get_hgvs_parser()
sv = hp.parse_hgvs_variant(hgvs_string)
ir = models.Identifier(namespace="NCBI", accession=sv.ac)
sequence_id = "VMC:GS_Ya6Rs7DHhDeg7YaOSg1EoNi3U_nQ9SvO" #get_vmc_sequence_id(ir)
if isinstance(sv.posedit.pos, hgvs.location.BaseOffsetInterval):
if sv.posedit.pos.start.is_intronic or sv.posedit.pos.end.is_intronic:
raise ValueError("Intronic HGVS variants are not supported".format(
sv.posedit.edit.type))
if sv.posedit.edit.type == 'ins':
interval = models.Interval(start=sv.posedit.pos.start.base,
end=sv.posedit.pos.start.base)
elif sv.posedit.edit.type in ('sub', 'del', 'delins'):
interval = models.Interval(start=sv.posedit.pos.start.base - 1,
end=sv.posedit.pos.end.base)
else:
raise ValueError("HGVS variant type {} is unsupported".format(
sv.posedit.edit.type))
location = models.Location(sequence_id=sequence_id, interval=interval)
location.id = computed_id(location)
state = sv.posedit.edit.alt or ''
allele = models.Allele(location_id=location.id, state=state)
allele.id = computed_id(allele)
bundle = models.Vmcbundle(
alleles={allele.id: allele.as_dict()},
genotypes={},
haplotypes={},
identifiers={sequence_id: [ir.as_dict()]},
locations={location.id: location.as_dict()},
meta={"version": "0.1"},
)
return ppj(bundle)
def add_hgvs_haplotype(self, hgvs_alleles, completeness="UNKNOWN"):
alleles = [self.add_hgvs_allele(hgvs_allele) for hgvs_allele in hgvs_alleles]
# create location from bounding box around alleles
sequence_ids = set(self.locations[a.location_id].sequence_id for a in alleles)
if len(sequence_ids) > 1:
raise Exception("Haplotypes must be defined on a single sequence")
sequence_id = next(iter(sequence_ids))
intervals = [self.locations[a.location_id].interval for a in alleles]
interval_min = min(int(i.start) for i in intervals)
interval_max = max(int(i.end) for i in intervals)
interval = models.Interval(start=interval_min, end=interval_max)
location = models.Location(sequence_id=sequence_id, interval=interval)
location.id = self._id_function(location)
self.locations[location.id] = location
haplotype = models.Haplotype(
completeness=completeness, location_id=location.id, allele_ids=[a.id for a in alleles])
haplotype.id = self._id_function(haplotype)
self.haplotypes[haplotype.id] = haplotype
return haplotype
def _make_vmc_allele(a):
"""given dict (from CAR json) for single genomicAllele or
transcriptAllele, create a (Location, Allele) pair, add to
the bundle, and return the allele.
"""
car_rsid = a["referenceSequence"].split("/")[-1]
ir = self._refseqmapper[car_rsid]
sequence_id = get_vmc_sequence_identifier(ir)
# N.B. Double check CA coordinate semantics
# If HGVS like re: insertions, then end -= 1 below
if len(a["coordinates"]) > 1:
_logger.warn(f"More than one coordinate set for resp[@id]; using only first")
coords = a["coordinates"][0]
interval = models.Interval(start=coords["start"] - 1, end=coords["end"])
location = models.Location(sequence_id=sequence_id, interval=interval)
location.id = computed_id(location)
allele = models.Allele(location_id=location.id, state=coords["allele"])
allele.id = computed_id(allele)
return (ir, sequence_id, location, allele)
import datetime
import json
from vmc import models, computed_id, serialize
# Interval
i = models.Interval(start=42, end=42)
assert "<Interval|42|42>" == serialize(i)
assert {"end": 42, "start": 42} == i.as_dict()
# Location
l = models.Location(sequence_id="VMC:GS_01234", interval=i)
assert "<Location|VMC:GS_01234|<Interval|42|42>>" == serialize(l)
l.id = computed_id(l)
assert "VMC:GL_OUqODzxryILUEDmv7uF8R8NwREJAx7gN" == l.id
assert {
"id": "VMC:GL_OUqODzxryILUEDmv7uF8R8NwREJAx7gN",
"interval": {
"end": 42,
"start": 42
},
"sequence_id": "VMC:GS_01234"
} == l.as_dict()
locations = {l.id: l.as_dict()}
# Allele
a = models.Allele(location_id=l.id, state="A")
assert "<Allele|VMC:GL_OUqODzxryILUEDmv7uF8R8NwREJAx7gN|A>" == serialize(a)
a.id = computed_id(a)
assert "VMC:GA_xTR0mmMviMLoAI9SwmDMFYr_AZczkjyU" == a.id
def build_loc(seq_id, start):
interval = models.Interval(start=start, end=start + 1)
location = models.Location(sequence_id=seq_id, interval=interval)
location.id = computed_id(location)
return location.id