BUG: account for genes being mapped directly to drug classes without …

…intermediate drugs in `confers_resistance_to()`. Whenever genes were mapped to drug classes (i.e. immediate child of antibiotic molecule), `confers_resistance_to()` would store these mappings in a temporary list (`backup_drugs`) and only use them if the ARG wasn't being mapped to any other drug. This strategy fails if an ARG is mapped to a drug class and a drug that falls under another drug class (not the same as the one being directly mapped to the ARG). In this case, the ARG is only mapped to the drug and the mapped drug class information is lost. Now, `confers_resistance_to()` does not utilize `backup_drugs`. Rather, it maps ARGs to every possible drug and drug class. It then iterates over these drugs and drug classes to check if any of the drugs are children nodes of mapped drug classes. Mapped drug classes that are the parents of mapped drugs are removed (they will be restored by `drugs_to_drug_classes()`).
BigDataBiology · Feb 5, 2025 · 8ebfb35 · 8ebfb35
1 parent 2b064e2
commit 8ebfb35
Show file tree

Hide file tree

Showing 13 changed files with 371 additions and 358 deletions.
diff --git a/argnorm/drug_categorization.py b/argnorm/drug_categorization.py
@@ -30,7 +30,7 @@ def _get_drug_classes(super_classes_list: List[str]) -> List[str]:
 
     return output
 
-def confers_resistance_to(aro_num: str) -> List[str]:
+def _get_drugs(aro_num: str) -> List[str]:
     '''
     Description: Returns a list of the drugs/antibiotics to which a gene confers resistance to.
 
@@ -41,34 +41,45 @@ def confers_resistance_to(aro_num: str) -> List[str]:
         target (list[str]):
             A list with ARO number of the drugs/antibiotics to which the input gene confers resistance to.
     '''
-    # 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 superclass in ARO[aro_num].superclasses():
         for drug in ARO[superclass.id].relationships.get(confers_resistance_to_drug_class_rel, []):
-            if list(ARO[drug.id].superclasses())[1:] == antibiotic_molecule_node:
-                backup_drugs.append(drug.id)
-            else:
-                target.add(drug.id)
+            target.add(drug.id)
 
         for drug in ARO[superclass.id].relationships.get(confers_resistance_to_antibiotic_rel, []):
-            if list(ARO[drug.id].superclasses())[1:] == antibiotic_molecule_node:
-                backup_drugs.append(drug.id)
-            else:
-                target.add(drug.id)
+            target.add(drug.id)
 
         for rel in [regulates_rel, participates_in_rel, part_of_rel]:
             for term in superclass.relationships.get(rel, []):
-                target.update(confers_resistance_to(term.id))
-
-    if not target:
-        target.update(backup_drugs)
+                target.update(_get_drugs(term.id))
 
     return sorted(target)
 
+def confers_resistance_to(aro_num: str) -> List[str]:
+    # 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
+
+    drugs = set(_get_drugs(aro_num))
+
+    drug_classes = set()
+    for drug in drugs:
+        if list(ARO[drug].superclasses())[1:] == antibiotic_molecule_node:
+            drug_classes.add(drug)
+
+    drugs = drugs - drug_classes
+    redundant_drug_classes = set()
+    for drug in drugs:
+        for drug_class in drug_classes:
+            if ARO[drug_class] in list(ARO[drug].superclasses())[1:]:
+                redundant_drug_classes.add(drug_class)
+
+    drug_classes = drug_classes - redundant_drug_classes
+    drugs.update(drug_classes)
+
+    return sorted(drugs)
+
 def drugs_to_drug_classes(drugs_list: List[str]) -> List[str]:
     '''
     Description: Returns a list of categories of drug classes, e.g. cephem and penam are categorized as beta_lactam antibiotics.

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.resfinder.tsv b/outputs/hamronized/abricate.resfinder.tsv
diff --git a/outputs/hamronized/abricate.resfinderfg.tsv b/outputs/hamronized/abricate.resfinderfg.tsv
@@ -3953,7 +3953,7 @@ Unnamed: 0	input_file_name	gene_symbol	gene_name	reference_database_id	reference
 3950	GMGC10.95nr_block_0294	D-alanyl-D-alanine	"D-alanyl-D-alanine carboxypeptidase DacB""|KU606669.1|Preterm_infant_stool|AMP"	resfinderfg	2021-Oct-18		abricate	abricate 1.0.1	84.47	GMGC10.269_887_786.DACB			598	1434					+		100.0											
 3951	GMGC10.95nr_block_0294	D-alanyl-D-alanine	"D-alanyl-D-alanine carboxypeptidase DacB""|KU606669.1|Preterm_infant_stool|AMP"	resfinderfg	2021-Oct-18		abricate	abricate 1.0.1	81.39	GMGC10.270_123_852.DACB			598	1430					+		99.52											
 3952	GMGC10.95nr_block_0294	D-alanyl-D-alanine	"D-alanyl-D-alanine carboxypeptidase DacB""|KU606669.1|Preterm_infant_stool|AMP"	resfinderfg	2021-Oct-18		abricate	abricate 1.0.1	81.12	GMGC10.280_419_486.DACB			598	1434					+		100.0											
-3953	GMGC10.95nr_block_0294	ABC-F	"ABC-F type ribosomal protection protein Msr(E)""|MG585948.1|pharmaceutical_effluent|AZM"	resfinderfg	2021-Oct-18		abricate	abricate 1.0.1	99.93	GMGC10.280_792_741.UNKNOWN			1	1476					+		100.0									ARO:3003109	ARO:0000006	ARO:0000000
+3953	GMGC10.95nr_block_0294	ABC-F	"ABC-F type ribosomal protection protein Msr(E)""|MG585948.1|pharmaceutical_effluent|AZM"	resfinderfg	2021-Oct-18		abricate	abricate 1.0.1	99.93	GMGC10.280_792_741.UNKNOWN			1	1476					+		100.0									ARO:3003109	ARO:0000006,ARO:0000026	ARO:0000000,ARO:0000026
 3954	GMGC10.95nr_block_0294	D-alanyl-D-alanine	"D-alanyl-D-alanine carboxypeptidase DacB""|KU606669.1|Preterm_infant_stool|AMP"	resfinderfg	2021-Oct-18		abricate	abricate 1.0.1	99.04	GMGC10.281_471_119.DACB			637	1473					+		100.0											
 3955	GMGC10.95nr_block_0294	D-alanyl-D-alanine	"D-alanyl-D-alanine carboxypeptidase DacB""|KU606669.1|Preterm_infant_stool|AMP"	resfinderfg	2021-Oct-18		abricate	abricate 1.0.1	80.05	GMGC10.281_560_329.DACB			598	1434					+		100.0											
 3956	GMGC10.95nr_block_0294	D-alanyl-D-alanine	"D-alanyl-D-alanine carboxypeptidase DacB""|KU606669.1|Preterm_infant_stool|AMP"	resfinderfg	2021-Oct-18		abricate	abricate 1.0.1	80.14	GMGC10.281_806_382.DACB			598	1433					+		99.88											

diff --git a/outputs/hamronized/amrfinderplus.ncbi.orfs.tsv b/outputs/hamronized/amrfinderplus.ncbi.orfs.tsv
@@ -18,6 +18,6 @@ amrfinderplus.ncbi.orfs.tsv	bexA	multidrug efflux MATE transporter BexA	NCBI Ref
 amrfinderplus.ncbi.orfs.tsv	aad9	ANT(9) family aminoglycoside nucleotidyltransferase	NCBI Reference Gene Database	2023-Nov-01	WP_002578722.1	amrfinderplus	3.10.30	gene_presence_detected	AMINOGLYCOSIDE	100.0			AMINOGLYCOSIDE		68	841	258			k119_82797										258				-	99.61	ARO:3002630	ARO:0000039	ARO:0000016
 amrfinderplus.ncbi.orfs.tsv	erm(B)	23S rRNA (adenine(2058)-N(6))-methyltransferase Erm(B)	NCBI Reference Gene Database	2023-Nov-01	WP_002292226.1	amrfinderplus	3.10.30	gene_presence_detected	MACROLIDE	100.0			MACROLIDE		24650	25384	245			k119_84636										245				-	100.0	ARO:3000375	ARO:0000006,ARO:0000027,ARO:0000046,ARO:0000057,ARO:0000065,ARO:0000066,ARO:3000145,ARO:3000156,ARO:3000158,ARO:3000176,ARO:3000583,ARO:3000584,ARO:3000669,ARO:3000672,ARO:3000673,ARO:3000674,ARO:3000675,ARO:3000677,ARO:3000678,ARO:3000679,ARO:3000680,ARO:3000681,ARO:3000682,ARO:3000867	ARO:0000000,ARO:0000000,ARO:0000000,ARO:0000000,ARO:0000000,ARO:0000000,ARO:0000000,ARO:0000000,ARO:0000000,ARO:0000017,ARO:0000017,ARO:0000026,ARO:0000026,ARO:0000026,ARO:0000026,ARO:0000026,ARO:0000026,ARO:0000026,ARO:0000026,ARO:0000026,ARO:0000026,ARO:0000026,ARO:0000026,ARO:0000026
 amrfinderplus.ncbi.orfs.tsv	lnu(AN2)	lincosamide nucleotidyltransferase Lnu(AN2)	NCBI Reference Gene Database	2023-Nov-01	WP_004308783.1	amrfinderplus	3.10.30	gene_presence_detected	LINCOSAMIDE	100.0			LINCOSAMIDE		1830	2339	170			k119_91457										170				-	100.0	ARO:3002835	ARO:0000046,ARO:3007169	ARO:0000017,ARO:0000017
-amrfinderplus.ncbi.orfs.tsv	mef(En2)	macrolide efflux MFS transporter Mef(En2)	NCBI Reference Gene Database	2023-Nov-01	WP_063853729.1	amrfinderplus	3.10.30	gene_presence_detected	MACROLIDE	100.0			MACROLIDE		2367	3569	401			k119_91457										401				-	99.5	ARO:3004659	ARO:0000046	ARO:0000017
+amrfinderplus.ncbi.orfs.tsv	mef(En2)	macrolide efflux MFS transporter Mef(En2)	NCBI Reference Gene Database	2023-Nov-01	WP_063853729.1	amrfinderplus	3.10.30	gene_presence_detected	MACROLIDE	100.0			MACROLIDE		2367	3569	401			k119_91457										401				-	99.5	ARO:3004659	ARO:0000000,ARO:0000046	ARO:0000000,ARO:0000017
 amrfinderplus.ncbi.orfs.tsv	tet(W)	tetracycline resistance ribosomal protection protein Tet(W)	NCBI Reference Gene Database	2023-Nov-01	WP_002586627.1	amrfinderplus	3.10.30	gene_presence_detected	TETRACYCLINE	100.0			TETRACYCLINE		21199	23115	639			k119_9485										639				+	100.0	ARO:3000194	ARO:0000051,ARO:0000069,ARO:3000152,ARO:3000528,ARO:3000667,ARO:3000668	ARO:3000050,ARO:3000050,ARO:3000050,ARO:3000050,ARO:3000050,ARO:3000050
 amrfinderplus.ncbi.orfs.tsv	catA13	type A-13 chloramphenicol O-acetyltransferase	NCBI Reference Gene Database	2023-Nov-01	WP_043774378.1	amrfinderplus	3.10.30	gene_presence_detected	CHLORAMPHENICOL	100.0			PHENICOL		1975	2595	207			k119_95290										207				+	100.0	ARO:3004454	ARO:3000385	ARO:3000387