ADD use all parents for gene to drug mapping

confers_resistance_to() now gets drugs for the whole AMR gene family. This is done by navigating up ARO tree and getting drugs for parent terms of input genes. For example, OXA-19 previously only returned cephalosporin and penam, but now will also return oxacillin (from AMR gene family).
BigDataBiology · Jun 23, 2024 · be22611 · be22611
1 parent d8ae8b0
commit be22611
Show file tree

Hide file tree

Showing 11 changed files with 548 additions and 530 deletions.
diff --git a/CHANGELOG.md b/CHANGELOG.md
@@ -2,6 +2,9 @@
 
 ## Unreleased
 
+### Update drug categorization
+- confers_resistance_to() now gets drugs for the whole AMR gene family. For example, OXA-19 previously only returned cephalosporin and penam, but now will also return oxacillin (from AMR gene family).
+
 ### Manual curation
 - argannot_curation: (Tet)tetH:EF460464:6286-7839:1554 was incorrectly annotated as ARO:3004797 which is a beta-lactamase due to a loose RGI hit. This was manually curated to ARO:3000175.
 

diff --git a/argnorm/drug_categorization.py b/argnorm/drug_categorization.py
@@ -7,27 +7,36 @@
 
 def confers_resistance_to(aro_num: str) -> List[str]:
     '''
-    Description: Returns a list of the drugs/antibiotics to which a gene confers resistance to.
+    Description: Returns a list of the drugs/antibiotics to which a gene and its parents confers resistance to.
 
     Parameters:
         aro_num (str): ARO number. Needs to be in the form 'ARO:number'.
 
     Returns:
         target (list[str]):
-            A list with ARO number of the drugs/antibiotics to which the input gene confers resistance to.
+            A list with ARO number of the drugs/antibiotics to which the input gene and its parents confers resistance to.
     '''
-
+    antibiotic_molecule_node = [ARO['ARO:1000003'], ARO['ARO:1000001']]
+    # some gene superclasses can map to drugs which are immediate children of 'antibiotic molecule'
+    # only use these if no other drugs can be found, as this information will be present in
+    # drugs to drug classes
+    backup_drugs = []
     target = set()
-
     for term in ARO[aro_num].superclasses():
         for drug in term.relationships.get(confers_resistance_to_drug_class_rel, []):
-            target.add(drug.id)
+            if list(ARO[drug.id].superclasses())[1:] == antibiotic_molecule_node:
+                backup_drugs.append(drug.id)
+            else:
+                target.add(drug.id)
 
         for drug in term.relationships.get(confers_resistance_to_antibiotic_rel, []):
-            target.add(drug.id)
+            if list(ARO[drug.id].superclasses())[1:] == antibiotic_molecule_node:
+                backup_drugs.append(drug.id)
+            else:
+                target.add(drug.id)
 
-        if target:
-            break
+    if not target:
+        target.update(backup_drugs)
 
     return sorted(target)
 

diff --git a/outputs/hamronized/abricate.argannot.tsv b/outputs/hamronized/abricate.argannot.tsv
diff --git a/outputs/hamronized/abricate.megares.tsv b/outputs/hamronized/abricate.megares.tsv
diff --git a/outputs/hamronized/abricate.ncbi.tsv b/outputs/hamronized/abricate.ncbi.tsv
diff --git a/outputs/hamronized/abricate.resfinder.tsv b/outputs/hamronized/abricate.resfinder.tsv
diff --git a/outputs/hamronized/abricate.resfinderfg.tsv b/outputs/hamronized/abricate.resfinderfg.tsv
diff --git a/outputs/hamronized/args-oap.sarg.reads.tsv b/outputs/hamronized/args-oap.sarg.reads.tsv
diff --git a/outputs/raw/args-oap.sarg.reads.tsv b/outputs/raw/args-oap.sarg.reads.tsv
diff --git a/outputs/raw/deeparg.deeparg.orfs.tsv b/outputs/raw/deeparg.deeparg.orfs.tsv
diff --git a/tests/test_arg_category.py b/tests/test_arg_category.py
@@ -7,13 +7,19 @@ def test_confers_resistance_to():
     test_cases = [
         "ARO:3003938",
         "ARO:3000014",
-        "ARO:3003725"
+        "ARO:3003725",
+        "ARO:3000230",
+        "ARO:3005046",
+        "ARO:3001414"
     ]
 
     expected_output = [
         ["ARO:3000008", "ARO:0000032", "ARO:0000044"],
         ["ARO:3000008", "ARO:0000004", "ARO:3003706", "ARO:0000032", "ARO:3000637"],
-        ["ARO:3000081"]
+        ["ARO:3000081"],
+        ["ARO:3007382", "ARO:0000052", "ARO:0000007", "ARO:0000035", "ARO:0000049"],
+        ["ARO:0000015", "ARO:3000008"],
+        ['ARO:0000032', 'ARO:0000056', 'ARO:3000008']
     ]
 
     for t,e in zip(test_cases, expected_output):