def generateVariantAnnotation(self, variant):
"""
Generate a random variant annotation based on a given variant.
This generator should be seeded with a value that is unique to the
variant so that the same annotation will always be produced regardless
of the order it is generated in.
"""
# To make this reproducible, make a seed based on this
# specific variant.
seed = self._randomSeed + variant.start + variant.end
randomNumberGenerator = random.Random()
randomNumberGenerator.seed(seed)
ann = protocol.VariantAnnotation()
ann.variantAnnotationSetId = str(self.getCompoundId())
ann.variantId = variant.id
ann.start = variant.start
ann.end = variant.end
ann.createDateTime = self._creationTime
# make a transcript effect for each alternate base element
# multiplied by a random integer (0,5)
ann.transcriptEffects = []
for base in variant.alternateBases * (
randomNumberGenerator.randint(0, 5)):
ann.transcriptEffects.append(self.generateTranscriptEffect(
ann, base, randomNumberGenerator))
ann.id = self.getVariantAnnotationId(variant, ann)
return ann
def testHashVariantAnnotation(self):
annotation = protocol.VariantAnnotation()
variant = protocol.Variant()
expected = 'bec63dc7c876bb3c7b71422203b101d1'
hashed = self._variantAnnotationSet.hashVariantAnnotation(
variant, annotation)
self.assertEqual(hashed, expected)
def testHashVariantAnnotation(self):
annotation = protocol.VariantAnnotation()
variant = protocol.Variant()
expected = hashlib.md5('\t()\t[]\t').hexdigest()
hashed = self._variantAnnotationSet.hashVariantAnnotation(
variant, annotation)
self.assertEqual(hashed, expected)
def _createGaVariantAnnotation(self):
"""
Convenience method to set the common fields in a GA VariantAnnotation
object from this variant set.
"""
ret = protocol.VariantAnnotation()
ret.created = self._creationTime
ret.updated = self._updatedTime
ret.variantAnnotationSetId = self.getId()
return ret
def testCreation(self):
dataset = datasets.AbstractDataset('dataset1')
localId = "variantAnnotationSetId"
simulatedVariantSet = variants.SimulatedVariantSet(
dataset, 'variantSet1', randomSeed=self.randomSeed,
numCalls=self.numCalls, variantDensity=self.variantDensity)
simulatedVariantAnnotationSet = variants.SimulatedVariantAnnotationSet(
dataset, localId, simulatedVariantSet, self.randomSeed)
annotations = simulatedVariantAnnotationSet.getVariantAnnotations(
self.referenceName, self.startPosition, self.endPosition)
self.assertEquals(
simulatedVariantSet.toProtocolElement().id,
simulatedVariantAnnotationSet.toProtocolElement().variantSetId,
"Variant Set ID should match the annotation's variant set ID")
for ann in annotations:
for key in protocol.VariantAnnotation().requiredFields:
self.assertEquals(datetime.datetime.strptime(
ann.createDateTime, "%Y-%m-%dT%H:%M:%S.%fZ").strftime(
"%Y-%m-%dT%H:%M:%S.%fZ"), ann.createDateTime,
"Expect time format to be in ISO8601")
self.assertTrue(hasattr(ann, key),
"Failed to find required key: " + key)