Add author_cell_type_annotations (JSON) per Cell_Cluster

pandasaurus_cxg/graph_generator/graph_generator.py

@@ -1,3 +1,4 @@
+import json
 import textwrap
 import uuid
 from enum import Enum
@@ -120,9 +121,7 @@ def generate_rdf_graph(self, merge: bool = False):
                         )
                     )
 
-            self.graph.add(
-                (dataset_class, URIRef(self.ns[ncname_safe(key)]), Literal(value))
-            )
+            self.graph.add((dataset_class, URIRef(self.ns[ncname_safe(key)]), Literal(value)))
         has_source = URIRef(HAS_SOURCE["iri"])
         self.graph.add((has_source, RDFS.label, Literal(HAS_SOURCE["label"])))
 
@@ -171,10 +170,25 @@ def generate_rdf_graph(self, merge: bool = False):
             resource = self.ns[_uuid]
             self.graph.add((resource, RDF.type, cell_set_class))
             self.graph.add((resource, has_source, dataset_class))
+            # Collect author_cell_type keys here
+            payload = {}
             for k, v in inner_dict.items():
                 if k in {"subcluster_of", "cluster_matches"}:
                     continue
-                self.graph.add((resource, self.ns[ncname_safe(k)], Literal(v)))
+                elif k in {"cell_count", "cell_type"}:
+                    self.graph.add((resource, self.ns[k], Literal(v)))
+                else:
+                    # Author annotations go into the JSON payload
+                    payload[k] = v
+            # Add one JSON literal for all author annotations
+            if payload:
+                self.graph.add(
+                    (
+                        resource,
+                        self.ns.author_cell_type_json,
+                        Literal(json.dumps(payload, ensure_ascii=False)),
+                    )
+                )
 
         # add relationship between each resource based on their predicate in the co_annotation_report
         subcluster = URIRef(SUBCLUSTER_OF.get("iri"))

pandasaurus_cxg/graph_generator/graph_generator_utils.py

@@ -147,8 +147,8 @@ def ncname_safe(term: str) -> str:
 
     """
     term = term.replace(" ", "_")
-    term = re.sub(r'^[^A-Za-z_]+', '', term)
-    return re.sub(r'[^A-Za-z0-9_\-\.]', '_', term)
+    term = re.sub(r"^[^A-Za-z_]+", "", term)
+    return re.sub(r"[^A-Za-z0-9_\-\.]", "_", term)
 
 
 def parse_citation_field_into_dict(value: str) -> Dict[str, str]:

test/graph_generator/test_graph_generator.py

@@ -199,7 +199,7 @@ def test_generate_rdf_graph_with_merge(graph_generator_instance_for_kidney, expe
         )
         == expected_stable_ids
     )
-    assert len(graph_generator.graph) == 747
+    assert len(graph_generator.graph) == 584
     assert (
         len([[s, p, o] for s, p, o in graph_generator.graph.triples((None, RDF.type, None))]) == 146
     )
@@ -216,7 +216,7 @@ def test_generate_rdf_graph_with_merge(graph_generator_instance_for_kidney, expe
                 )
             ]
         )
-        == 90
+        == 77
     )
     assert (
         len(
@@ -241,7 +241,7 @@ def test_generate_rdf_graph_with_merge(graph_generator_instance_for_kidney, expe
 def test_generate_rdf_graph_without_merge(graph_generator_instance_for_kidney):
     graph_generator = graph_generator_instance_for_kidney
     graph_generator.generate_rdf_graph()
-    assert len(graph_generator.graph) == 2177
+    assert len(graph_generator.graph) == 1398
     assert (
         len([[s, p, o] for s, p, o in graph_generator.graph.triples((None, RDF.type, None))]) == 312
     )
@@ -270,11 +270,11 @@ def test_enrich_rdf_graph_with_merge(graph_generator_instance_for_kidney):
     graph_generator = graph_generator_instance_for_kidney
     graph_generator.generate_rdf_graph(merge=True)
 
-    assert len(graph_generator.graph) == 747
+    assert len(graph_generator.graph) == 584
 
     graph_generator.enrich_rdf_graph()
 
-    assert len(graph_generator.graph) == 1242
+    assert len(graph_generator.graph) == 1081
     assert (
         URIRef(CONSIST_OF.get("iri")),
         RDFS.label,
@@ -291,19 +291,19 @@ def test_enrich_rdf_graph_with_merge(graph_generator_instance_for_kidney):
                 if str(s).startswith("http://purl.obolibrary.org/obo/CL_")
             ]
         )
-        == 531
+        == 529
     )
 
 
 def test_enrich_rdf_graph_without_merge(graph_generator_instance_for_kidney):
     graph_generator = graph_generator_instance_for_kidney
     graph_generator.generate_rdf_graph()
 
-    assert len(graph_generator.graph) == 2177
+    assert len(graph_generator.graph) == 1398
 
     graph_generator.enrich_rdf_graph()
 
-    assert len(graph_generator.graph) == 2674
+    assert len(graph_generator.graph) == 1895
 
 
 def test_save_rdf_graph(graph_generator_instance_for_kidney):