Refactor such that Uniprot IDs are node properties in xrefs column (i…

…nstead of edges)
Knowledge-Graph-Hub · Jun 17, 2020 · cf1672d · cf1672d
1 parent e388719
commit cf1672d
Show file tree

Hide file tree

Showing 3 changed files with 27 additions and 30 deletions.
diff --git a/kg_covid_19/transform_utils/string_ppi/node_header.json b/kg_covid_19/transform_utils/string_ppi/node_header.json
@@ -4,5 +4,6 @@
     "category",
     "description",
     "alias",
+    "xrefs",
     "provided_by"
-]
+]
diff --git a/kg_covid_19/transform_utils/string_ppi/string_ppi.py b/kg_covid_19/transform_utils/string_ppi/string_ppi.py
@@ -18,16 +18,15 @@
 Write node and edge headers that look something like:
 
 Node: 
-id  name    category
-protein:1234    TBX4    Protein 
+id  name    category    xrefs   provided_by
+protein:1234    TBX4    biolink:Protein UniprotKB:123456    STRING 
+
+xrefs contains the UniprotKB id for the protein, if available
 
 Edge: 
 subject edge_label  object  relation
 protein:1234    interacts_with  protein:4567    RO:0002434
 
-Also write edges that create xrefs between ENSP ids and UniprotKB IDs, like so:
-subject edge_label  object  relation
-protein:1234    bl:xrefs  protein:4567    RO:0002434
 
 """
 
@@ -188,6 +187,7 @@ def run(self, data_file: Optional[str] = None) -> None:
                                     'biolink:Gene',
                                     gene_informations['description'],
                                     f"NCBIGene:{self.ensembl2ncbi_map[gene]}",
+                                    "",
                                     self.source_name
                                 ]
                             )
@@ -206,32 +206,25 @@ def run(self, data_file: Optional[str] = None) -> None:
                     # write node data
                     if protein not in seen_proteins:
                         seen_proteins.add(protein)
-                        write_node_edge_item(
-                            fh=node,
-                            header=self.node_header,
-                            data=[f"ENSEMBL:{protein}", "",
-                                  protein_node_type, "", "", self.source_name]
-                        )
 
                         # if we have an equivalent Uniprot ID for this Ensembl protein
                         # ID make an xref edge, and a node for the Uniprot ID
+                        uniprot_curie = ''
                         if protein in string_to_uniprot_id_map:
                             uniprot_curie = \
                                 f"UniprotKB:{string_to_uniprot_id_map[protein]}"
-                            write_node_edge_item(
-                                fh=node,
-                                header=self.node_header,
-                                data=[uniprot_curie, "",
-                                      protein_node_type, "", "", self.source_name])
-                            write_node_edge_item(
-                                fh=edge,
-                                header=self.edge_header,
-                                data=[f"ENSEMBL:{protein}",
-                                      "biolink:xrefs",
-                                      uniprot_curie,
-                                      "biolink:xrefs",
-                                      "uniprot",
-                                      ] + self.extra_header)
+
+                        write_node_edge_item(
+                            fh=node,
+                            header=self.node_header,
+                            data=[f"ENSEMBL:{protein}", "",
+                                  protein_node_type,
+                                  "",
+                                  "",
+                                  uniprot_curie,  # xref
+                                  self.source_name]
+                        )
+
 
                 # write edge data
                 write_node_edge_item(

diff --git a/tests/test_string.py b/tests/test_string.py
@@ -56,23 +56,26 @@ def test_nodes_file(self):
         node_file = os.path.join(self.string_output_dir, "nodes.tsv")
         self.assertTrue(os.path.isfile(node_file))
         node_df = pd.read_csv(node_file, sep="\t", header=0)
-        self.assertEqual((11, 6), node_df.shape)
-        self.assertEqual(['id', 'name', 'category', 'description', 'alias',
+        self.assertEqual((10, 7), node_df.shape)
+        self.assertEqual(['id', 'name', 'category', 'description', 'alias', 'xrefs',
                           'provided_by'], list(node_df.columns))
-        self.assertCountEqual(['UniprotKB:P84085', 'ENSEMBL:ENSP00000000233',
+        self.assertCountEqual(['ENSEMBL:ENSP00000000233',
                               'ENSEMBL:ENSP00000272298', 'ENSEMBL:ENSP00000253401',
                               'ENSEMBL:ENSP00000401445', 'ENSEMBL:ENSP00000418915',
                               'ENSEMBL:ENSP00000327801', 'ENSEMBL:ENSP00000466298',
                               'ENSEMBL:ENSP00000232564', 'ENSEMBL:ENSP00000393379',
                               'ENSEMBL:ENSP00000371253'],
                              list(node_df.id.unique()))
+        self.assertCountEqual('UniprotKB:P84085',
+                              node_df.loc[node_df['id'] ==
+                                          'ENSEMBL:ENSP00000000233'].xrefs.item())
 
     def test_edges_file(self):
         self.string.run()
         edge_file = os.path.join(self.string_output_dir, "edges.tsv")
         self.assertTrue(os.path.isfile(edge_file))
         edge_df = pd.read_csv(edge_file, sep="\t", header=0)
-        self.assertEqual((10, 19), edge_df.shape)
+        self.assertEqual((9, 19), edge_df.shape)
         self.assertEqual(['subject', 'edge_label', 'object', 'relation', 'provided_by',
                           'combined_score', 'neighborhood', 'neighborhood_transferred',
                           'fusion', 'cooccurence', 'homology', 'coexpression',