Merge pull request #216 from Knowledge-Graph-Hub/add_example_sparql_q…

…ueries

Refactor query command to execute SPARQL queries described in YAML files

kg_covid_19/query.py

@@ -1,14 +1,36 @@
 import logging
 
-from kg_covid_19.query_utils.target_candidates.target_candidates \
-    import TargetCandidates
+import yaml
+from SPARQLWrapper import SPARQLWrapper, JSON, XML, TURTLE, N3, RDF, RDFXML, CSV, TSV  # type: ignore
 
-QUERIES = {
-    'TargetCandidates': TargetCandidates
-}
 
+def run_query(query: str, endpoint: str, return_format=JSON) -> dict:
+    sparql = SPARQLWrapper(endpoint)
+    sparql.setQuery(query)
+    sparql.setReturnFormat(return_format)
+    results = sparql.query().convert()
 
-def run_query(query: str, input_dir: str, output_dir: str) -> None:
-    logging.info(f"Running query {query}")
-    t = QUERIES[query](input_dir, output_dir)
-    t.run()
+    return results
+
+
+def parse_query_yaml(yaml_file) -> dict:
+    return yaml.load(open(yaml_file))
+
+
+def result_dict_to_tsv(result_dict: dict, outfile: str) -> None:
+    with open(outfile, 'wt') as f:
+        # header
+        f.write("\t".join(result_dict['head']['vars']) + "\n")
+        for row in result_dict['results']['bindings']:
+            row_items = []
+            for col in result_dict['head']['vars']:
+                try:
+                    row_items.append(row[col]['value'])
+                except KeyError:
+                    logging.error('Problem retrieving result for col %s in row %s' %
+                                  (col, "\t".join(row)))
+                    row_items.append('ERROR')
+            try:
+                f.write("\t".join(row_items) + "\n")
+            except:
+                pass

queries/sparql/query-01-bl-cat-counts.yaml

@@ -0,0 +1,11 @@
+title:
+  "Get counts for biolink categories"
+description:
+  "Get counts for biolink categories"
+endpoint:
+  "http://kg-hub-rdf.berkeleybop.io/blazegraph/sparql"
+query: >
+  SELECT (COUNT(?v2) AS ?v1) ?v0
+  WHERE {
+    ?v2 <https://w3id.org/biolink/vocab/category> ?v0
+  } GROUP BY ?v0

queries/sparql/query-02-sars-cov-2-prot.yaml

@@ -0,0 +1,17 @@
+title:
+  "Get SARS-CoV-2 proteins"
+description:
+  "Get info about SARS-CoV-2 proteins in the graph"
+endpoint:
+  "http://kg-hub-rdf.berkeleybop.io/blazegraph/sparql"
+query: >
+  prefix owl: <http://www.w3.org/2002/07/owl#>
+  prefix rdfs: <http://www.w3.org/2000/01/rdf-schema#>
+  prefix rdf: <http://www.w3.org/1999/02/22-rdf-syntax-ns#>
+  prefix bl: <https://w3id.org/biolink/vocab/>
+  SELECT ?protein ?p ?v WHERE
+  {
+  ?protein bl:category bl:Protein; <https://www.example.org/UNKNOWN/ncbi_taxid> ?ncbitaxonid;
+   ?p ?v .
+   FILTER(?ncbitaxonid="2697049")
+  }

queries/sparql/query-03-sars-cov-2-interactors.yaml

@@ -0,0 +1,25 @@
+title:
+  "Get human proteins that interact with SARS-CoV-2 proteins"
+description:
+  "Get human proteins that interact with SARS-CoV-2 proteins"
+endpoint:
+  "http://kg-hub-rdf.berkeleybop.io/blazegraph/sparql"
+query: >
+  prefix owl: <http://www.w3.org/2002/07/owl#>
+  prefix rdfs: <http://www.w3.org/2000/01/rdf-schema#>
+  prefix rdf: <http://www.w3.org/1999/02/22-rdf-syntax-ns#>
+  prefix bl: <https://w3id.org/biolink/vocab/>
+  SELECT ?covp ?covplab ?p ?plab ?humanp ?humanplab WHERE
+  {
+  ?interaction bl:edge_label bl:interacts_with .
+  ?interaction bl:object ?covp .
+  ?interaction bl:subject ?humanp .
+  ?interaction bl:relation ?p .
+  ?covp bl:category bl:Protein; <https://www.example.org/UNKNOWN/ncbi_taxid> ?covtaxon .
+  ?humanp bl:category bl:Protein; <https://www.example.org/UNKNOWN/ncbi_taxid> ?humantaxon .
+  OPTIONAL { ?covp rdfs:label ?covplab } .
+  OPTIONAL { ?humanp rdfs:label ?humanplab } .
+  OPTIONAL { ?p rdfs:label ?plab }
+   FILTER(?covtaxon="2697049")
+   FILTER(?humantaxon="9606")
+  }