Merge pull request #475 from biolink/issue-470-meta-property-propagation

Issue#470 - fix propagation of 'deprecate' metadata flag

kgx/source/obograph_source.py

@@ -1,5 +1,4 @@
 import gzip
-import tarfile
 import typing
 from itertools import chain
 from typing import Optional, Tuple, Dict, Generator, Any
@@ -104,6 +103,9 @@ def read_node(self, node: Dict) -> Optional[Tuple[str, Dict]]:
         curie = self.prefix_manager.contract(node["id"])
         node_properties = {}
         if "meta" in node:
+            # Returns a dictionary that contains 'description', 'subsets',
+            # 'synonym', 'xrefs', a 'deprecated' flag and/or
+            # 'equivalent_nodes', if the corresponding key values are set
             node_properties = self.parse_meta(node["id"], node["meta"])
 
         fixed_node = dict()
@@ -114,22 +116,30 @@ def read_node(self, node: Dict) -> Optional[Tuple[str, Dict]]:
 
         if "description" in node_properties:
             fixed_node["description"] = node_properties["description"]
+
+        if "subsets" in node_properties:
+            fixed_node["subsets"] = node_properties["subsets"]
+
         if "synonym" in node_properties:
             fixed_node["synonym"] = node_properties["synonym"]
+
         if "xrefs" in node_properties:
             fixed_node["xref"] = node_properties["xrefs"]
-        if "subsets" in node_properties:
-            fixed_node["subsets"] = node_properties["subsets"]
+
+        if "deprecated" in node_properties:
+            fixed_node["deprecated"] = node_properties["deprecated"]
 
         if "category" not in node:
             category = self.get_category(curie, node)
             if category:
                 fixed_node["category"] = [category]
             else:
                 fixed_node["category"] = ["biolink:OntologyClass"]
+
         if "equivalent_nodes" in node_properties:
             equivalent_nodes = node_properties["equivalent_nodes"]
             fixed_node["same_as"] = equivalent_nodes
+
         return super().read_node(fixed_node)
 
     def read_edges(self, filename: str, compression: Optional[str] = None) -> Generator:
@@ -301,8 +311,8 @@ def parse_meta(self, node: str, meta: Dict) -> Dict:
         Returns
         -------
         Dict
-            A dictionary that contains 'description', 'synonyms',
-            'xrefs', and 'equivalent_nodes'.
+            A dictionary that contains 'description', 'subsets',
+            'synonyms', 'xrefs', a 'deprecated' flag and/or 'equivalent_nodes'.
 
         """
         # cross species links are in meta; this needs to be parsed properly too

tests/resources/phenio.json

@@ -0,0 +1,53 @@
+{
+  "graphs": [
+    {
+      "nodes": [
+        {
+          "id": "http://purl.obolibrary.org/obo/GO_0051370",
+          "lbl": "obsolete ZASP binding",
+          "type": "INDIVIDUAL",
+          "meta": {
+            "definition": {
+              "val": "OBSOLETE. Binding to Z-band alternatively spliced PDZ motif protein (ZASP). ZASP is a Z-band protein specifically expressed in heart and skeletal muscle. This protein contains N-terminal PDZ domain and C-terminal LIM domain.",
+              "xrefs": [
+                "PMID:10427098",
+                "PMID:11699871"
+              ]
+            },
+            "comments": [
+              "This term was made obsolete because it represents binding to an individual protein."
+            ],
+            "synonyms": [
+              {
+                "pred": "hasExactSynonym",
+                "val": "Z-band alternatively spliced PDZ-motif protein binding"
+              },
+              {
+                "pred": "hasExactSynonym",
+                "val": "ZASP binding"
+              }
+            ],
+            "basicPropertyValues": [
+              {
+                "pred": "http://purl.obolibrary.org/obo/IAO_0100001",
+                "val": "GO:0008092"
+              },
+              {
+                "pred": "http://www.geneontology.org/formats/oboInOwl#hasOBONamespace",
+                "val": "molecular_function"
+              }
+            ],
+            "deprecated": true
+          }
+        }
+      ],
+      "edges": [],
+      "id": "http://purl.obolibrary.org/obo/phenio.owl",
+      "meta": {
+        "subsets": [],
+        "xrefs": [],
+        "basicPropertyValues": []
+      }
+    }
+  ]
+}

tests/unit/test_source/test_obograph_source.py

@@ -2,9 +2,10 @@
 
 import pytest
 
-from kgx.source import ObographSource
+from kgx.cli import transform
+from kgx.source import ObographSource, TsvSource
 from kgx.transformer import Transformer
-from tests import RESOURCE_DIR
+from tests import RESOURCE_DIR, TARGET_DIR
 
 
 def test_read_obograph1():
@@ -107,6 +108,52 @@ def test_read_jsonl2():
     assert "GO slim generic" in n2["provided_by"]
 
 
+def test_read_deprecated_term():
+    """
+    Read from an PATO JSON using ObographSource,
+    to validate capture of "deprecate" status
+    """
+    t = Transformer()
+    s = ObographSource(t)
+    g = s.parse(
+        os.path.join(RESOURCE_DIR, "pato.json"),
+        knowledge_source="Phenotype and Trait Ontology",
+    )
+    nodes = {}
+    for rec in g:
+        if rec:
+            if len(rec) != 4:
+                nodes[rec[0]] = rec[1]
+
+    n1 = nodes["PATO:0000000"]
+    assert n1["id"] == "PATO:0000000"
+    assert n1["name"] == "obsolete pato"
+    assert n1["deprecated"] is True
+
+
+def test_read_deprecated_term_phenio():
+    """
+    Read from a Phenio JSON using ObographSource,
+    to validate capture of "deprecate" status
+    """
+    t = Transformer()
+    s = ObographSource(t)
+    g = s.parse(
+        os.path.join(RESOURCE_DIR, "phenio.json"),
+        knowledge_source="Phenomics Integrative Ontology",
+    )
+    nodes = {}
+    for rec in g:
+        if rec:
+            if len(rec) != 4:
+                nodes[rec[0]] = rec[1]
+
+    n1 = nodes["GO:0051370"]
+    assert n1["id"] == "GO:0051370"
+    assert n1["name"] == "obsolete ZASP binding"
+    assert n1["deprecated"] is True
+
+
 @pytest.mark.parametrize(
     "query",
     [
@@ -205,3 +252,34 @@ def test_error_detection():
         t.write_report(None, "Error")
     if len(t.get_errors("Warning")) > 0:
         t.write_report(None, "Warning")
+
+
+def test_phenio_obojson_to_tsv():
+    """
+    Testing transitive propagation of node properties
+    (mainly node 'deprecated' status)
+    from a Phenio JSON to TSV file format
+    """
+    transform(
+        inputs=[os.path.join(RESOURCE_DIR, "phenio.json")],
+        input_format="obojson",
+        output=os.path.join(TARGET_DIR, "phenio"),
+        output_format="tsv",
+        stream=False
+    )
+
+    tin = Transformer()
+    s = TsvSource(tin)
+
+    g = s.parse(filename=os.path.join(TARGET_DIR, "phenio_nodes.tsv"), format="tsv")
+
+    nodes = {}
+    for rec in g:
+        if rec:
+            if len(rec) != 4:
+                nodes[rec[0]] = rec[1]
+
+    n1 = nodes["GO:0051370"]
+    assert n1["id"] == "GO:0051370"
+    assert n1["name"] == "obsolete ZASP binding"
+    assert n1["deprecated"] is True