Attribute mapping from node to dug element (#203)

* Release/2.8.0 (#198) * Bumping version * support for extracting dug elements from graph (#197) * support for extracting dug elements from graph * adding flag for enabling dug element extraction from graph * adding new config for node_to dug element parsing * adding more parameters to crawler to able configuration to element extraction logic * add tests * add tests for crawler Co-authored-by: Yaphetkg <[email protected]> * Update _version.py * Update _version.py updating version for final push to master * Update factory.py Adding more comments Co-authored-by: Carl Schreep <[email protected]> Co-authored-by: Yaphetkg <[email protected]> * Release/v2.9.0 (#201) * Bumping version * support for extracting dug elements from graph (#197) * support for extracting dug elements from graph * adding flag for enabling dug element extraction from graph * adding new config for node_to dug element parsing * adding more parameters to crawler to able configuration to element extraction logic * add tests * add tests for crawler Co-authored-by: Yaphetkg <[email protected]> * Display es scores (#199) * Include ES scores in variable results * Round ES score to 6 * Update _version.py (#200) * Update _version.py Co-authored-by: Carl Schreep <[email protected]> Co-authored-by: Yaphetkg <[email protected]> Co-authored-by: Ginnie Hench <[email protected]> * adding more config options for node extraction * some refactoring Co-authored-by: Carl Schreep <[email protected]> Co-authored-by: Yaphetkg <[email protected]> Co-authored-by: Ginnie Hench <[email protected]>
helxplatform · Apr 1, 2022 · 256d1e0 · 256d1e0
1 parent 4c9dbc3
commit 256d1e0
Show file tree

Hide file tree

Showing 5 changed files with 64 additions and 19 deletions.
diff --git a/src/dug/config.py b/src/dug/config.py
@@ -67,7 +67,15 @@ class Config:
         # Dug element type to cast the query kg nodes to
         "cde": {
             # Parse nodes matching criteria in kg
-            "node_type": "biolink:Publication"
+            "node_type": "biolink:Publication",
+            "curie_prefix": "HEALCDE",
+            "attribute_mapping": {
+                # "DugElement Attribute" : "KG Node attribute"
+                "name": "name",
+                "desc": "summary",
+                "collection_name": "cde_category",
+                "collection_id":  "cde_category"
+            }
         }
     })
 

diff --git a/src/dug/core/crawler.py b/src/dug/core/crawler.py
@@ -206,6 +206,8 @@ def expand_to_dug_element(self,
         elements = []
         # using node_type as the primary criteria for matching nodes to element type.
         target_node_type = casting_config["node_type"]
+        curie_filter = casting_config["curie_prefix"]
+        attribute_mapping = casting_config["attribute_mapping"]
         target_node_type_snake_case = biolink_snake_case(target_node_type.replace("biolink:", ""))
         for ident_id, identifier in concept.identifiers.items():
 
@@ -244,14 +246,14 @@ def expand_to_dug_element(self,
                     # and return the variables.
                     for node_id, node in answer.nodes.items():
                         if target_node_type in node["category"]:
-                            # @TODO make element creation more generic
-                            # @TODO need to encode more data into the graph nodes, to parse them properly
-                            element = DugElement(
-                                elem_id=node_id,
-                                name=node.get('name', ""),
-                                desc=node.get('summary', ""),
-                                elem_type=dug_element_type
-                            )
-                            element.add_concept(concept)
-                            elements.append(element)
+                            if node['id'].startswith(curie_filter):
+                                element_attribute_args = {"elem_id": node_id, "elem_type": dug_element_type}
+                                element_attribute_args.update({key: node.get(attribute_mapping[key], "")
+                                                               for key in attribute_mapping
+                                                               })
+                                element = DugElement(
+                                    **element_attribute_args
+                                )
+                                element.add_concept(concept)
+                                elements.append(element)
         return elements
diff --git a/src/dug/core/factory.py b/src/dug/core/factory.py
@@ -96,7 +96,9 @@ def build_element_extraction_parameters(self, source=None):
             {
                 "output_dug_type": dug_type,
                 "casting_config": {
-                    "node_type": queries[dug_type]['node_type']
+                    "node_type": queries[dug_type]["node_type"],
+                    "curie_prefix": queries[dug_type]["curie_prefix"],
+                    "attribute_mapping": queries[dug_type]["attribute_mapping"]
                     # CDE's are only ones
                     # but if we had two biolink:Publication nodes we want to conditionally
                     # cast to other output_dug_type, we could extend this config

diff --git a/tests/unit/mocks/data/tranql_response.json b/tests/unit/mocks/data/tranql_response.json
@@ -8,14 +8,14 @@
         }
       },
       "nodes": {
+        "publication": {
+          "category": "biolink:Publication"
+        },
         "disease": {
           "category": "biolink:Disease",
           "id": [
             "MONDO:0008187"
           ]
-        },
-        "publication": {
-          "category": "biolink:Publication"
         }
       }
     },
@@ -25,11 +25,11 @@
           "name": "panic disorder 1",
           "category": [
             "biolink:Disease",
-            "biolink:Entity",
+            "biolink:DiseaseOrPhenotypicFeature",
             "biolink:ThingWithTaxon",
             "biolink:BiologicalEntity",
-            "biolink:NamedThing",
-            "biolink:DiseaseOrPhenotypicFeature"
+            "biolink:Entity",
+            "biolink:NamedThing"
           ],
           "attributes": [
             {
@@ -72,6 +72,30 @@
               "value": "Filename: Photosensitivity_PAQ_CDE_v1.0.json; File path: Supplemental Questionnaires/Sensory/Photosensitivity Assessment Questionnaire (PAQ); Photosensitivity Assessment Questionnaire",
               "name": "summary"
             },
+            {
+              "type": "NA",
+              "value": [
+                "Supplemental Questionnaires",
+                "Sensory",
+                "Photosensitivity Assessment Questionnaire (PAQ)"
+              ],
+              "name": "cde_categories"
+            },
+            {
+              "type": "NA",
+              "value": [
+                "Supplemental Questionnaires",
+                "Adult/Pediatric",
+                "Acute/Chronic Pain",
+                "Photosensitivity Assessment Questionnaire (PAQ)"
+              ],
+              "name": "cde_category_extended"
+            },
+            {
+              "type": "NA",
+              "value": "Supplemental Questionnaires",
+              "name": "cde_category"
+            },
             {
               "type": "NA",
               "value": [

diff --git a/tests/unit/test_crawler.py b/tests/unit/test_crawler.py
@@ -84,7 +84,16 @@ def test_expand_to_dug_element(crawler):
     concept.add_identifier(identifier)
     new_elements = crawler.expand_to_dug_element(
         concept=concept,
-        casting_config={"node_type": "biolink:Publication"},
+        casting_config={
+            "node_type": "biolink:Publication",
+            "curie_prefix": "HEALCDE",
+            "attribute_mapping": {
+                "name": "name",
+                "desc": "summary",
+                "collection_name": "cde_category",
+                "collection_id":  "cde_category"
+            }
+        },
         dug_element_type="test-element",
         tranql_source="test:graph"
     )