Add transitivity function in pandas_utils

kg_bacdive/transform.py

@@ -4,7 +4,7 @@
 from typing import List, Optional
 
 from kg_bacdive.transform_utils.bacdive.bacdive import BacDiveTransform
-from kg_bacdive.transform_utils.mediadive.mediadive import MediaDiveDiveTransform
+from kg_bacdive.transform_utils.mediadive.mediadive import MediaDiveTransform
 from kg_bacdive.transform_utils.ontology.ontology_transform import ONTOLOGIES, OntologyTransform
 from kg_bacdive.transform_utils.traits.traits import TraitsTransform
 
@@ -19,7 +19,7 @@
     # "ProteinAtlasTransform": ProteinAtlasTransform,
     # "STRINGTransform": STRINGTransform,
     "BacDiveTransform": BacDiveTransform,
-    "MediaDiveDiveTransform": MediaDiveDiveTransform,
+    "MediaDiveTransform": MediaDiveTransform,
     "TraitsTransform": TraitsTransform,
 }
 

kg_bacdive/transform_utils/constants.py

@@ -68,7 +68,7 @@
 CHEBI_PREFIX = "CHEBI:"
 CAS_RN_PREFIX = "CAS-RN:"
 PUBCHEM_PREFIX = "PubChem:"
-MEDIADIVE_COMPOUND_PREFIX = "mediadive.ingredient:"
+MEDIADIVE_INGREDIENT_PREFIX = "mediadive.ingredient:"
 MEDIADIVE_SOLUTION_PREFIX = "mediadive.solution:"
 MEDIADIVE_MEDIUM_PREFIX = "mediadive.medium:"
 GO_PREFIX = "GO:"

kg_bacdive/transform_utils/mediadive/__init__.py

@@ -1,4 +1,4 @@
 """BacDive transform."""
-from .mediadive import MediaDiveDiveTransform
+from .mediadive import MediaDiveTransform
 
-__all__ = ["MediaDiveDiveTransform"]
+__all__ = ["MediaDiveTransform"]

kg_bacdive/transform_utils/mediadive/mediadive.py

@@ -43,9 +43,9 @@
     KEGG_KEY,
     KEGG_PREFIX,
     MEDIADIVE_COMPLEX_MEDIUM_COLUMN,
-    MEDIADIVE_COMPOUND_PREFIX,
     MEDIADIVE_DESC_COLUMN,
     MEDIADIVE_ID_COLUMN,
+    MEDIADIVE_INGREDIENT_PREFIX,
     MEDIADIVE_LINK_COLUMN,
     MEDIADIVE_MAX_PH_COLUMN,
     MEDIADIVE_MEDIUM_PREFIX,
@@ -72,10 +72,10 @@
     SOLUTIONS_KEY,
 )
 from kg_bacdive.transform_utils.transform import Transform
-from kg_bacdive.utils.pandas_utils import drop_duplicates
+from kg_bacdive.utils.pandas_utils import drop_duplicates, establish_transitive_relationship
 
 
-class MediaDiveDiveTransform(Transform):
+class MediaDiveTransform(Transform):
 
     """Template for how the transform class would be designed."""
 
@@ -144,7 +144,7 @@ def standardize_compound_id(self, id: str):
         elif data[CAS_RN_KEY] is not None:
             return CAS_RN_PREFIX + str(data[CAS_RN_KEY])
         else:
-            return MEDIADIVE_COMPOUND_PREFIX + id
+            return MEDIADIVE_INGREDIENT_PREFIX + id
 
     def download_yaml_and_get_json(
         self,
@@ -296,4 +296,10 @@ def run(self, data_file: Union[Optional[Path], Optional[str]] = None):
                     progress.update()
 
         drop_duplicates(self.output_node_file)
-        drop_duplicates(self.output_edge_file)
+        establish_transitive_relationship(
+            self.output_edge_file,
+            MEDIADIVE_MEDIUM_PREFIX,
+            MEDIADIVE_SOLUTION_PREFIX,
+            MEDIUM_TO_INGREDIENT_EDGE,
+            MEDIADIVE_INGREDIENT_PREFIX,
+        )

kg_bacdive/utils/pandas_utils.py

@@ -3,6 +3,8 @@
 
 import pandas as pd
 
+from kg_bacdive.transform_utils.constants import OBJECT_COLUMN, PREDICATE_COLUMN, SUBJECT_COLUMN
+
 
 def drop_duplicates(file_path: Path):
     """
@@ -14,3 +16,55 @@ def drop_duplicates(file_path: Path):
     df = pd.read_csv(file_path, sep="\t", low_memory=False)
     df = df.drop_duplicates()
     df.to_csv(file_path, sep="\t", index=False)
+    return df
+
+
+def establish_transitive_relationship(
+    file_path: Path,
+    subject_prefix: str,
+    intermediate_prefix: str,
+    predicate: str,
+    object_prefix: str,
+) -> pd.DataFrame:
+    """
+    Establish transitive relationship given the predicate is the same.
+
+    e.g.: Existent relations:
+        1. A => predicate => B
+        2. B => predicate => C
+
+    This function adds the relation A => predicate => C
+
+
+    :param file_path: Filepath of the edge file.
+    :param subject_prefix: Subject prefix (A in the example)
+    :param intermediate_prefix: Intermediate prefix that connects the subject to object (B in the example).
+    :param predicate: The common predicate between all relations.
+    :param object_prefix: Object prefix (C in the example)
+    :return: Core dataframe with additional deduced rows.
+    """
+    df = drop_duplicates(file_path)
+    df_relations = df.loc[df[PREDICATE_COLUMN] == predicate]
+    subject_condition = df_relations[SUBJECT_COLUMN].str.startswith(subject_prefix)
+    intermediate_subject_condition = df_relations[SUBJECT_COLUMN].str.startswith(
+        intermediate_prefix
+    )
+    object_condition = df_relations[OBJECT_COLUMN].str.startswith(object_prefix)
+    intermediate_object_condition = df_relations[OBJECT_COLUMN].str.startswith(intermediate_prefix)
+    subject_intermediate_df = df_relations[subject_condition & intermediate_object_condition]
+    intermediate_object_df = df_relations[intermediate_subject_condition & object_condition]
+
+    list_of_dfs_to_append = []
+
+    for row in subject_intermediate_df.iterrows():
+        transitive_relations_df = intermediate_object_df.loc[
+            intermediate_object_df[SUBJECT_COLUMN] == row[1].object
+        ]
+        transitive_relations_df.loc[
+            transitive_relations_df[SUBJECT_COLUMN] == row[1].object, SUBJECT_COLUMN
+        ] = row[1].subject
+        list_of_dfs_to_append.append(transitive_relations_df)
+
+    df = pd.concat([df] + list_of_dfs_to_append)
+    df.to_csv(file_path, sep="\t", index=False)
+    return df

merged_graph_stats.yaml

@@ -9,13 +9,13 @@ edge_stats:
         unknown:
           count: 36244
     biolink:has_part:
-      count: 84816
+      count: 102140
       provided_by:
         unknown:
-          count: 84816
+          count: 102140
       source:
         unknown:
-          count: 84816
+          count: 102140
     biolink:has_phenotype:
       count: 19727
       provided_by:
@@ -124,13 +124,13 @@ edge_stats:
         unknown:
           count: 2027
     biolink:ChemicalEntity-biolink:has_part-biolink:ChemicalEntity:
-      count: 84816
+      count: 102140
       provided_by:
         unknown:
-          count: 84816
+          count: 102140
       source:
         unknown:
-          count: 84816
+          count: 102140
     biolink:ChemicalEntity-biolink:has_part-biolink:ChemicalSubstance:
       count: 50254
       provided_by:
@@ -464,7 +464,7 @@ edge_stats:
   - unknown
   source:
   - unknown
-  total_edges: 1324137
+  total_edges: 1341461
 graph_name: kg-bacdive graph
 node_stats:
   count_by_category: