Merge pull request #97 from geneontology/issue-94-get-models-by-taxon

refactor SPARQL to be faster and just return models by taxon id
geneontology · Jun 18, 2024 · ee56256 · ee56256
2 parents 9411c34 + b8554bf
commit ee56256
Show file tree

Hide file tree

Showing 30 changed files with 97 additions and 295 deletions.
diff --git a/.github/workflows/pr-test-qc.yaml b/.github/workflows/pr-test-qc.yaml
@@ -2,14 +2,16 @@ name: go-fastapi-QC
 
 on:
   pull_request:
+    branches:
+      - no_op
     types: [opened, synchronize, reopened]
 
 jobs:
   build:
     runs-on: ubuntu-latest
     strategy:
       matrix:
-        python-version: [ "3.9", "3.10" ]
+        python-version: [ "3.11", "3.12" ]
 
     steps:
       - uses: actions/[email protected]

diff --git a/app/main.py b/app/main.py
@@ -1,4 +1,5 @@
 """main application entry point."""
+
 import uvicorn
 from fastapi import FastAPI
 from fastapi.middleware.cors import CORSMiddleware

diff --git a/app/middleware/logging_middleware.py b/app/middleware/logging_middleware.py
@@ -1,4 +1,5 @@
 """Middleware to log requests."""
+
 import logging
 
 from fastapi import Request
@@ -12,7 +13,6 @@
 
 
 class LoggingMiddleware(BaseHTTPMiddleware):
-
     """Middleware to log requests."""
 
     async def dispatch(self, request: Request, call_next):

diff --git a/app/routers/bioentity.py b/app/routers/bioentity.py
@@ -1,4 +1,5 @@
 """bioentity router."""
+
 import logging
 from enum import Enum
 from typing import List
@@ -29,7 +30,6 @@
 
 
 class RelationshipType(str, Enum):
-
     """
     Enumeration for Gene Ontology relationship types used for filtering associations.
 

diff --git a/app/routers/labeler.py b/app/routers/labeler.py
@@ -1,4 +1,5 @@
 """Provides a route for fetching labels for CURIEs/IDs."""
+
 import logging
 from typing import List
 

diff --git a/app/routers/models.py b/app/routers/models.py
@@ -621,28 +621,23 @@ async def get_term_details_by_taxon_id(
         final_taxon = final_taxon + new_taxon
     query = (
         """
-        PREFIX metago: <http://model.geneontology.org/>
-        PREFIX dc: <http://purl.org/dc/elements/1.1/>
-        PREFIX enabled_by: <http://purl.obolibrary.org/obo/RO_0002333>
-        PREFIX in_taxon: <http://purl.obolibrary.org/obo/RO_0002162>
+            PREFIX metago: <http://model.geneontology.org/>
+            PREFIX dc: <http://purl.org/dc/elements/1.1/>
+            PREFIX enabled_by: <http://purl.obolibrary.org/obo/RO_0002333>
+            PREFIX in_taxon: <http://purl.obolibrary.org/obo/RO_0002162>
+            PREFIX pav: <http://purl.org/pav/>
+            PREFIX prov: <http://www.w3.org/ns/prov#>
+            PREFIX xsd: <http://www.w3.org/2001/XMLSchema#>
+            PREFIX biolink: <https://w3id.org/biolink/vocab/>
 
-        SELECT distinct ?gocam
+            SELECT ?gocam
 
-        WHERE
-        {
-            GRAPH ?gocam {
-                ?gocam metago:graphType metago:noctuaCam .
-                ?s enabled_by: ?gpnode .
-                ?gpnode rdf:type ?identifier .
-                FILTER(?identifier != owl:NamedIndividual) .
-            }
-
-            ?identifier rdfs:subClassOf ?v0 .
-            ?identifier rdfs:label ?name .
+            WHERE
+                {
+                    ?gocam rdf:type owl:Ontology ;
+                    biolink:in_taxon <%s> ;
 
-            ?v0 owl:onProperty in_taxon: .
-            ?v0 owl:someValuesFrom <%s>
-        }
+                }
     """
         % final_taxon
     )

diff --git a/app/routers/ontology.py b/app/routers/ontology.py
@@ -1,4 +1,5 @@
 """Ontology-related endpoints."""
+
 import json
 import logging
 from enum import Enum
@@ -22,7 +23,6 @@
 
 
 class GraphType(str, Enum):
-
     """Enum for the different types of graphs that can be returned."""
 
     topology_graph = "topology_graph"

diff --git a/app/routers/pathway_widget.py b/app/routers/pathway_widget.py
@@ -1,4 +1,5 @@
 """Pathway widget API endpoints."""
+
 import logging
 
 from curies import Converter

diff --git a/app/routers/prefixes.py b/app/routers/prefixes.py
@@ -1,4 +1,5 @@
 """Module contains the API endpoints for handling prefixes and expansions."""
+
 import logging
 
 from curies import Converter

diff --git a/app/routers/search.py b/app/routers/search.py
@@ -1,4 +1,5 @@
 """search router."""
+
 import logging
 from enum import Enum
 
@@ -14,7 +15,6 @@
 
 
 class AutocompleteCategory(str, Enum):
-
     """The category of items to retrieve, can be 'gene' or 'term'."""
 
     gene = "gene"

diff --git a/app/routers/slimmer.py b/app/routers/slimmer.py
@@ -1,4 +1,5 @@
 """slimmer router."""
+
 import logging
 from enum import Enum
 from typing import List
@@ -20,7 +21,6 @@
 
 
 class RelationshipType(str, Enum):
-
     """Relationship type for slimmer."""
 
     acts_upstream_of_or_within = ACTS_UPSTREAM_OF_OR_WITHIN

diff --git a/app/routers/users_and_groups.py b/app/routers/users_and_groups.py
@@ -1,4 +1,5 @@
 """The users and groups endpoints."""
+
 import logging
 
 from fastapi import APIRouter, Path
@@ -77,9 +78,7 @@ async def get_user_by_orcid(
         PREFIX MF: <http://purl.obolibrary.org/obo/GO_0003674>
         PREFIX CC: <http://purl.obolibrary.org/obo/GO_0005575>
 
-		SELECT  ?name  			(GROUP_CONCAT(distinct ?organization;separator="@|@") AS ?organizations)
- 	                       		(GROUP_CONCAT(distinct ?affiliationIRI;separator="@|@") AS ?affiliationsIRI)
-                        		(GROUP_CONCAT(distinct ?affiliation;separator="@|@") AS ?affiliations)
+		SELECT  ?name
 								(GROUP_CONCAT(distinct ?gocam;separator="@|@") as ?gocams)
 								(GROUP_CONCAT(distinct ?date;separator="@|@") as ?gocamsDate)
 								(GROUP_CONCAT(distinct ?title;separator="@|@") as ?gocamsTitle)
@@ -126,15 +125,6 @@ async def get_user_by_orcid(
 
   			?goid rdfs:label ?goname .
 
-            # Getting some information on the contributor
-            optional { ?orcidIRI rdfs:label ?name } .
-            BIND(IF(bound(?name), ?name, ?orcid) as ?name) .
-            optional { ?orcidIRI vcard:organization-name ?organization } .
-            optional {
-                ?orcidIRI has_affiliation: ?affiliationIRI .
-                ?affiliationIRI rdfs:label ?affiliation
-            } .
-
             # crash the query for SYNGO user "http://orcid.org/0000-0002-1190-4481"^^xsd:string
             optional {
   			?gpid rdfs:label ?gpname .
@@ -150,9 +140,6 @@ async def get_user_by_orcid(
     collated = {}
     results = si._sparql_query(query)
     for result in results:
-        collated["organizations"] = result["organizations"].get("value")
-        collated["affiliations"] = result["affiliations"].get("value")
-        collated["affiliationsIRI"] = result["affiliationsIRI"].get("value")
         collated["gocams"] = result["gocams"].get("value")
         collated["gocamsDate"] = result["gocamsDate"].get("value")
         collated["gocamsTitle"] = result["gocamsTitle"].get("value")

diff --git a/app/utils/ontology_utils.py b/app/utils/ontology_utils.py
@@ -1,4 +1,5 @@
 """ontology utility functions."""
+
 import logging
 
 from linkml_runtime.utils.namespaces import Namespaces

diff --git a/app/utils/prefix_utils.py b/app/utils/prefix_utils.py
@@ -1,4 +1,5 @@
 """prefix utility functions."""
+
 import logging
 
 from curies import Converter

diff --git a/app/utils/settings.py b/app/utils/settings.py
@@ -1,4 +1,5 @@
 """settings for the application."""
+
 import logging
 from enum import Enum
 from os import path
@@ -41,21 +42,18 @@ def get_golr_config():
 
 
 class ESOLR(Enum):
-
     """Enum for the GOLR URL."""
 
     GOLR = get_golr_config()["solr_url"]["url"]
 
 
 class ESPARQL(Enum):
-
     """Enum for the SPARQL endpoint URL."""
 
     SPARQL = get_sparql_endpoint()
 
 
 class ESOLRDoc(Enum):
-
     """Enum for the GOLR document type."""
 
     ONTOLOGY = "ontology_class"