Interaction participants identifier bug

PathwayMerger · Oct 4, 2018 · 672cb95 · 672cb95
1 parent 8eeb548
commit 672cb95
Show file tree

Hide file tree

Showing 2 changed files with 14 additions and 5 deletions.
diff --git a/src/pathme/wikipathways/rdf_sparql.py b/src/pathme/wikipathways/rdf_sparql.py
@@ -83,7 +83,7 @@
        ?uri_id rdf:type ?uri_type .
        ?uri_id wp:participants ?participants_entry .
        optional {?participants_entry dcterms:identifier ?participants_id .}
-       optional {?participants_entry dcterms:bdbEntrezGene ?ncbigene_participants .}
+       optional {?participants_entry wp:bdbEntrezGene ?ncbigene_participants .}
     }
     """
 
@@ -100,8 +100,8 @@
        ?uri_id wp:target ?target_entry .
        optional {?source_entry dcterms:identifier ?dc_source .}
        optional {?target_entry dcterms:identifier ?dc_target .}
-       optional {?source_entry dcterms:bdbEntrezGene ?ncbigene_source .}
-       optional {?target_entry dcterms:bdbEntrezGene ?ncbigene_target .}
+       optional {?source_entry wp:bdbEntrezGene ?ncbigene_source .}
+       optional {?target_entry wp:bdbEntrezGene ?ncbigene_target .}
     }
     """
 

diff --git a/src/pathme/wikipathways/utils.py b/src/pathme/wikipathways/utils.py
@@ -113,12 +113,12 @@ def get_valid_gene_identifier(node_ids_dict, hgnc_manager):
         return _validate_query(hgnc_manager, hgnc_entry, ensembl_id, ENSEMBL)
 
     elif 'ec-code' in node_ids_dict['uri_id']:
-        ec_number = node_ids_dict['name']
+        ec_number = check_multiple(node_ids_dict['name'], 'ec-code')
         return EXPASY, ec_number, ec_number
 
     # Only wikipathways identifier is given
     elif 'bdb_wikidata' in node_ids_dict:
-        name = node_ids_dict['name']
+        name = check_multiple(node_ids_dict['name'], 'wikidata')
 
         # Find out whether the name is a valid HGNC symbol
         hgnc_entry = hgnc_manager.get_gene_by_hgnc_symbol(name)
@@ -130,6 +130,15 @@ def get_valid_gene_identifier(node_ids_dict, hgnc_manager):
         log.warning('Adding WikiPathways node %s (%s)', name, WIKIPATHWAYS)
         return WIKIPATHWAYS, name, name
 
+    elif 'wikipedia' in node_ids_dict['uri_id']:
+        wiki_name = check_multiple(node_ids_dict['identifier'], 'wikipedia_id')
+        wiki_id = check_multiple(node_ids_dict['name'], 'wikipedia_name')
+        return 'WIKIPEDIA', wiki_name, wiki_id
+
+    elif 'identifier' and 'namespace' in node_ids_dict:
+        node_id_dict = {'bdb_'+node_ids_dict['namespace']: node_ids_dict['identifier']}
+        return get_valid_gene_identifier(node_id_dict, hgnc_manager)
+
     raise Exception('Unknown identifier for node %s', node_ids_dict)