Merge pull request #196 from gyorilab/background_set

Expose background gene list option in analysis functions

src/indra_cogex/analysis/gene_analysis.py

@@ -38,6 +38,7 @@ def discrete_analysis(
         minimum_evidence_count: int = 1,
         minimum_belief: float = 0,
         indra_path_analysis: bool = False,
+        background_gene_list: List[str] = None,
         *,
         client: Neo4jClient
 ) -> Dict[str, Union[pd.DataFrame, None]]:
@@ -59,6 +60,10 @@ def discrete_analysis(
         Minimum belief score for filtering, by default 0.
     indra_path_analysis : bool, optional
         Whether to perform INDRA pathway analysis, by default False.
+    background_gene_list : List[str], optional
+        A list of background genes of which the gene list is a part
+        to constrain the space of possible genes to consider
+        when calculating enrichment statistics.
     client : Neo4jClient, optional
         The Neo4j client, managed automatically by the autoclient decorator.
 
@@ -74,6 +79,12 @@ def discrete_analysis(
             f"Failed to parse the following gene identifiers: {', '.join(errors)}"
         )
 
+    if background_gene_list:
+        background_genes, _ = parse_gene_list(background_gene_list)
+        background_gene_ids = list(background_genes)
+    else:
+        background_gene_ids = None
+
     results = {}
     for analysis_name, analysis_func in [
         ("go", go_ora),
@@ -87,7 +98,8 @@ def discrete_analysis(
         if analysis_name in {"go", "wikipathways", "reactome", "phenotype"}:
             analysis_result = analysis_func(
                 client=client, gene_ids=gene_set, method=method, alpha=alpha,
-                keep_insignificant=keep_insignificant
+                keep_insignificant=keep_insignificant,
+                background_gene_ids=background_gene_ids
             )
         else:
             # Run INDRA analysis if enabled
@@ -96,7 +108,8 @@ def discrete_analysis(
                     client=client, gene_ids=gene_set, method=method, alpha=alpha,
                     keep_insignificant=keep_insignificant,
                     minimum_evidence_count=minimum_evidence_count,
-                    minimum_belief=minimum_belief
+                    minimum_belief=minimum_belief,
+                    background_gene_ids=background_gene_ids
                 )
             else:
                 continue

src/indra_cogex/client/enrichment/discrete.py

@@ -215,7 +215,9 @@ def go_ora(
         if not background_gene_ids
         else len(background_gene_ids)
     )
-    return _do_ora(get_go(client=client), query=gene_ids, count=count, **kwargs)
+    bg_genes = frozenset(background_gene_ids) if background_gene_ids else None
+    return _do_ora(get_go(client=client, background_gene_ids=bg_genes),
+                   query=gene_ids, count=count, **kwargs)
 
 
 def wikipathways_ora(
@@ -249,8 +251,10 @@ def wikipathways_ora(
         if not background_gene_ids
         else len(background_gene_ids)
     )
+    bg_genes = frozenset(background_gene_ids) if background_gene_ids else None
     return _do_ora(
-        get_wikipathways(client=client), query=gene_ids, count=count, **kwargs
+        get_wikipathways(client=client, background_gene_ids=bg_genes),
+        query=gene_ids, count=count, **kwargs
     )
 
 
@@ -285,7 +289,9 @@ def reactome_ora(
         if not background_gene_ids
         else len(background_gene_ids)
     )
-    return _do_ora(get_reactome(client=client), query=gene_ids, count=count, **kwargs)
+    bg_genes = frozenset(background_gene_ids) if background_gene_ids else None
+    return _do_ora(get_reactome(client=client, background_gene_ids=bg_genes),
+                   query=gene_ids, count=count, **kwargs)
 
 
 @autoclient()
@@ -321,8 +327,10 @@ def phenotype_ora(
         if not background_gene_ids
         else len(background_gene_ids)
     )
+    bg_genes = frozenset(background_gene_ids) if background_gene_ids else None
     return _do_ora(
-        get_phenotype_gene_sets(client=client), query=gene_ids, count=count, **kwargs
+        get_phenotype_gene_sets(client=client, background_gene_ids=bg_genes),
+        query=gene_ids, count=count, **kwargs
     )
 
 
@@ -367,11 +375,13 @@ def indra_downstream_ora(
         if not background_gene_ids
         else len(background_gene_ids)
     )
+    bg_genes = frozenset(background_gene_ids) if background_gene_ids else None
     return _do_ora(
         get_entity_to_regulators(
             client=client,
             minimum_evidence_count=minimum_evidence_count,
             minimum_belief=minimum_belief,
+            background_gene_ids=bg_genes
         ),
         query=gene_ids,
         count=count,
@@ -420,11 +430,13 @@ def indra_upstream_ora(
         if not background_gene_ids
         else len(background_gene_ids)
     )
+    bg_genes = frozenset(background_gene_ids) if background_gene_ids else None
     return _do_ora(
         get_entity_to_targets(
             client=client,
             minimum_evidence_count=minimum_evidence_count,
             minimum_belief=minimum_belief,
+            background_gene_ids=bg_genes
         ),
         query=gene_ids,
         count=count,