def main(args):
##
qmods = {
'extraConstraint':
makeOneOfConstraint('Gene.primaryIdentifier', args.identifiers),
}
##
qs = [
('gene', mouseGenes),
('synonyms', mouseSynonyms),
('expressed', mouseExpressedGenes),
('expressedImages', mouseExpressedGenesWithImages),
('location', mouseLocations),
('proteinIds', mouseProteinIds),
('xrefs', mouseXrefs),
('pantherId', mousePantherIds),
('myGeneLink', mouseMyGeneLinks),
]
hasPheno = set()
for r in doQuery(mouseHasPheno, MOUSEMINE):
hasPheno.add(r['primaryIdentifier'])
hasImpc = set()
for r in doQuery(mouseHasImpc, MOUSEMINE):
hasImpc.add(r['primaryIdentifier'])
id2gene = {}
for label, q in qs:
if label == 'gene':
for r in doQuery(q % qmods, MOUSEMINE):
r['mgiid'] = r['primaryIdentifier']
id2gene[r['primaryIdentifier']] = r
else:
for r in doQuery(q % qmods, MOUSEMINE):
obj = id2gene.get(r['primaryIdentifier'], None)
if obj:
obj.setdefault(label, []).append(r)
print('{\n "metaData": %s,\n "data": [' %
json.dumps(buildMetaObject(MOUSEMINE), indent=2))
first = True
for i in id2gene:
obj = id2gene[i]
if not first: print(',', end='')
obj["hasPheno"] = obj["primaryIdentifier"] in hasPheno
obj["hasImpc"] = obj["primaryIdentifier"] in hasImpc
print(json.dumps(getJsonObj(obj), indent=2))
first = False
print(']\n}')
def getSamples(url):
eid2samples = {}
for r in doQuery(htSamples, url):
rkey = (r["samples.name"], r["samples.age"], r["samples.sex"],
r["samples.structure.identifier"])
eid2samples.setdefault(r['experimentId'], {})[rkey] = r
return eid2samples
def main():
args = parseCmdLine()
ids = args.identifiers
xtra = makeOneOfConstraint('Genotype.alleles.feature.primaryIdentifier',
ids)
xtra2 = makeOneOfConstraint('Allele.feature.primaryIdentifier', ids)
# Process Genotype-AllelePairs. Build index from genotype id to list of component (allele+state)
id2components = {}
toDelete = set(SKIP)
for r in doQuery(q_genotypeAlleles % xtra, MOUSEMINE):
gid = r["primaryIdentifier"]
try:
id2components.setdefault(gid, []).append(r)
except:
toDelete.add(gid)
# Build set of MGI ids of alleles being sent to the alliance
includedAlleles = set()
for r in doQuery(q_alleles % xtra2, MOUSEMINE):
includedAlleles.add(r['primaryIdentifier'])
# Process genotypes. For each one, find / attach its components if any and output.
# Screen for genotypes to be deleted.
#
print('{\n "metaData": %s,\n "data": [' %
json.dumps(buildMetaObject(MOUSEMINE), indent=2))
first = True
for g in doQuery(q_genotypes % xtra, MOUSEMINE):
gid = g["primaryIdentifier"]
if gid in toDelete:
continue
g["components"] = id2components.get(gid, [])
gobj = getJsonObj(g, includedAlleles)
if gobj:
if not first: print(",", end=' ')
print(json.dumps(gobj, indent=2))
first = False
print("]}")
def loadEMAPAParents(url):
log('Loading EMAPA parents...')
q = '''<query
name=""
model="genomic"
view="OntologyRelation.childTerm.identifier
OntologyRelation.parentTerm.identifier"
longDescription=""
sortOrder="OntologyRelation.childTerm.identifier asc"
>
<constraint path="OntologyRelation.childTerm" type="EMAPATerm"/>
<constraint path="OntologyRelation.parentTerm" type="EMAPATerm"/>
<constraint path="OntologyRelation.direct" op="=" value="true"/>
</query>'''
id2pids = {}
for i, r in enumerate(doQuery(q, url)):
id2pids.setdefault(r["childTerm.identifier"],
[]).append(r["parentTerm.identifier"])
log('Loaded %d parent/child relations.' % i)
return id2pids
def loadEMAPA(url):
log('Loading EMAPA...')
id2emapa = {}
q = '''
<query
model="genomic"
view="
EMAPATerm.identifier
EMAPATerm.name
EMAPATerm.startsAt
EMAPATerm.endsAt
"
>
</query>
'''
for t in doQuery(q, url):
t["startsAt"] = int(t["startsAt"])
t["endsAt"] = int(t["endsAt"])
id2emapa[t["identifier"]] = t
log('Loaded %d EMAPA terms.' % len(id2emapa))
return id2emapa
def getExpressionData(url, ids):
log('Getting expression data...')
q = '''<query
model="genomic"
view="
GXDExpression.assayId
GXDExpression.assayType
GXDExpression.feature.primaryIdentifier
GXDExpression.stage
GXDExpression.structure.identifier
GXDExpression.publication.mgiId
GXDExpression.publication.pubMedId"
sortOrder="GXDExpression.assayId asc GXDExpression.structure.identifier asc GXDExpression.stage asc"
constraintLogic="A and (B or (C and D)) and E"
>
<constraint path="GXDExpression.detected" code="A" op="=" value="true"/>
<constraint path="GXDExpression.genotype.hasMutantAllele" code="B" op="=" value="false"/>
<constraint path="GXDExpression.assayType" code="C" op="=" value="In situ reporter (knock in)"/>
<constraint path="GXDExpression.genotype.zygosity" code="D" op="=" value="ht"/>
%s
</query>
''' % makeOneOfConstraint('GXDExpression.feature.primaryIdentifier', ids)
prev = None
qcount = 0
ycount = 0
for r in doQuery(q, MOUSEMINE):
qcount += 1
if not prev \
or r["assayId"] != prev["assayId"] \
or r["stage"] != prev["stage"] \
or r["structure.identifier"] != prev["structure.identifier"]:
ycount += 1
yield r
#
prev = r
#
log('getExpressionData: %d results => %d unique results' %
(qcount, ycount))
def loadSubmittedAlleles () :
aids = set()
for a in doQuery(qAlleles, MOUSEMINE):
aids.add(a["primaryIdentifier"])
return aids
def getExperiments(url):
for r in doQuery(htExperiments, url):
yield r
def getVariables(url):
eid2vars = {}
for r in doQuery(htVariables, url):
if r["variables.name"]:
eid2vars.setdefault(r['experimentId'], []).append(r)
return eid2vars
def getReferences(url):
eid2refs = {}
for r in doQuery(htReferences, url):
eid2refs.setdefault(r['experimentId'], []).append(r)
return eid2refs
def annotations(url, okind, skind, ids=None):
qopts = {
'alleleFeatView':
"OntologyAnnotation.subject.feature.primaryIdentifier"
if skind == "Allele" else "",
'xtraConstraint':
makeOneOfConstraint("OntologyAnnotation.subject.primaryIdentifier",
ids),
'xtraConstraint2':
makeOneOfConstraint(
"OntologyAnnotationEvidence.annotation.subject.primaryIdentifier",
ids),
'okind':
okind,
'skind':
skind,
}
qAnnots = '''<query
model="genomic"
view="
OntologyAnnotation.id
OntologyAnnotation.subject.primaryIdentifier
OntologyAnnotation.subject.symbol
OntologyAnnotation.subject.name
OntologyAnnotation.ontologyTerm.identifier
OntologyAnnotation.ontologyTerm.name
OntologyAnnotation.qualifier
%(alleleFeatView)s
"
sortOrder="OntologyAnnotation.id asc"
>
<constraint path="OntologyAnnotation.ontologyTerm" type="%(okind)s"/>
<constraint path="OntologyAnnotation.subject" type="%(skind)s"/>
<constraint path="OntologyAnnotation.subject.organism.taxonId" op="=" value="10090"/>
%(xtraConstraint)s
</query>
''' % qopts
qEvidence = '''<query
model="genomic"
view="
OntologyAnnotation.id
OntologyAnnotation.evidence.id
OntologyAnnotation.evidence.annotationDate
OntologyAnnotation.evidence.code.code
OntologyAnnotation.evidence.publications.id
OntologyAnnotation.evidence.publications.pubMedId
OntologyAnnotation.evidence.publications.mgiJnum
OntologyAnnotation.evidence.publications.mgiId
%(alleleFeatView)s
"
sortOrder="OntologyAnnotation.id asc OntologyAnnotation.evidence.id asc"
>
<constraint path="OntologyAnnotation.ontologyTerm" type="%(okind)s"/>
<constraint path="OntologyAnnotation.subject" type="%(skind)s"/>
<constraint path="OntologyAnnotation.subject.organism.taxonId" op="=" value="10090"/>
%(xtraConstraint)s
</query>
''' % qopts
qBaseAnnots = '''<query
model="genomic"
view="
OntologyAnnotationEvidence.id
OntologyAnnotationEvidence.publications.id
OntologyAnnotationEvidence.annotation.id
OntologyAnnotationEvidence.baseAnnotations.subject.primaryIdentifier
OntologyAnnotationEvidence.baseAnnotations.evidence.annotationDate
OntologyAnnotationEvidence.baseAnnotations.evidence.publications.pubMedId
"
sortOrder="OntologyAnnotationEvidence.annotation.id asc OntologyAnnotationEvidence.id asc"
>
<constraint path="OntologyAnnotationEvidence.annotation.ontologyTerm" type="%(okind)s"/>
<constraint path="OntologyAnnotationEvidence.annotation.subject" type="%(skind)s"/>
<constraint path="OntologyAnnotationEvidence.baseAnnotations.evidence.publications" op="=" loopPath="OntologyAnnotationEvidence.publications"/>
%(xtraConstraint2)s
</query>
''' % qopts
qs = [
map(lambda x: (x[0], 'annotation', list(x[1])),
groupby(doQuery(qAnnots, url), lambda e: e['id'])),
map(lambda x: (x[0], 'evidence', list(x[1])),
groupby(doQuery(qEvidence, url), lambda e: e['id'])),
map(lambda x: (x[0], 'baseAnnots', list(x[1])),
groupby(doQuery(qBaseAnnots, url), lambda e: e['annotation.id'])),
]
for x in groupby(heapq.merge(*qs), lambda x: x[0]):
r = {}
for y in x[1]:
if y[1] == 'annotation':
r.update(y[2][0])
elif y[1] == 'evidence':
r['evidence'] = y[2]
elif y[1] == 'baseAnnots':
# Note that these are *all* the base annotations. Each one is associated with a
# specific evidence object. Look for matching ids
# the field named 'id' in the base annot record should equal the 'evidence.id'
# in the evidence object.
r['baseAnnots'] = y[2]
rr = applyConversions(r, okind, skind)
if rr:
for n, e in enumerate(rr["invevidence"]):
rr["agrevidence"] = e
rr["agrbaseannots"] = rr["invbaseannots"][n]
rr2 = formatDafJsonRecord(
rr, "disease" if okind == "DOTerm" else "phenotype", skind)
if rr2:
yield rr2
def getAlleles(url, ids):
qopts = {
'xtraConstraint':
makeOneOfConstraint('Allele.feature.primaryIdentifier', ids)
}
# Query for alleles that have expressed component
qexpressors = '''<query
model="genomic"
view="
MGIExpressesComponent.allele.primaryIdentifier
MGIExpressesComponent.allele.symbol
"
></query>
'''
expressors = set()
for r in doQuery(qexpressors, url):
expressors.add(r['allele.primaryIdentifier'])
# Query allele synonyms, build index of id -> synonyms
# 2020-12-18: change constraints:
# - drop ontology annotation requirement
# - drop null allele type restriction
# - add exclusion when germline transmission = 'cell line'
qsynonyms = '''<query
model="genomic"
view="
Allele.primaryIdentifier
Allele.synonyms.value
"
constraintLogic="A and B and C and D"
sortOrder="Allele.primaryIdentifier asc"
>
<constraint code="A" path="Allele.organism.taxonId" op="=" value="10090" />
<constraint code="B" path="Allele.alleleType" op="NONE OF">
<value>QTL</value>
</constraint>
<constraint code="C" path="Allele.isWildType" op="=" value="false" />
<constraint code="D" path="Allele.glTransmission" op="!=" value="Cell Line"/>
%(xtraConstraint)s
</query>
''' % qopts
aid2syns = {}
for r in doQuery(qsynonyms, url):
aid2syns.setdefault(r['primaryIdentifier'],
set()).add(r['synonyms.value'])
# Main allele query.
qalleles = '''<query
model="genomic"
view="
Allele.primaryIdentifier
Allele.symbol
Allele.name
Allele.alleleType
Allele.molecularNote
Allele.feature.primaryIdentifier
Allele.feature.mgiType
Allele.drivenBy
"
constraintLogic="A and B and C and (D or E)"
sortOrder="Allele.primaryIdentifier asc"
>
<constraint code="A" path="Allele.organism.taxonId" op="=" value="10090" />
<constraint code="B" path="Allele.alleleType" op="NONE OF">
<value>QTL</value>
</constraint>
<constraint code="C" path="Allele.isWildType" op="=" value="false" />
<constraint code="D" path="Allele.glTransmission" op="!=" value="Cell Line"/>
<constraint code="E" path="Allele.glTransmission" op="IS NULL" />
%(xtraConstraint)s
</query>
''' % qopts
for r in doQuery(qalleles, url):
aid = r['primaryIdentifier']
r['synonyms'] = list(aid2syns.get(aid, []))
# If the allele has a driver or has expressed components, then the allele has a
# "construct". At the Alliance, constructs must have an ID, but at MGI they don't (they're not objects).
# So we create a fake ID for it. This are not displayed and are not used to create links.
# Constructs are dumped separately. Here we just need the ID
if r['drivenBy'] or aid in expressors:
r['construct'] = aid + "_con"
#
yield r