ENH: check RGI mapping accuracy and manually curate incorrect hits

All RGI hits are checked for mapping accuracy using `check_mapping_accuracy.py`. This is done by comparing the drug categorization of the ARO assigned to the ARG with the drug categorization assigned to the ARG by its original database. A list of ARGs with mismatched mappings are included in `manual_curation/`. Mismatched hits are manually curated to correct ARGs and manual curation files are updated. Some mismatched hits are marked `correct` or `incorrect`. `correct` hits have a drug category mismatch but have an ARO mapping that's determined to be correct. `incorrect` hits have a drug category mismatch and a manually curated ARO can't be found for them. Both `correct` and `incorrect` hits should not be included in the final manual curation. For megares, a file called `megares_meta_biocide_and_virulence_genes.tsv` is also created with metal, biocide, and virulence genes. These should not be mapped to any ARO terms and can be manually curated to no ARO. Metal, biocide, and virulence genes are directly added to manual curation files for other databases. Groot is derived from resfinder and argannot.

A list of all mismatched hits, with their manual curation can be found in `db_harmonisation/all_mismatched_hits.tsv`

argnorm/data/groot_ARO_mapping.tsv

@@ -1512,7 +1512,7 @@ Original ID	ARO
 (Gly)VanA-Ao2:HQ679900:1034-2080:1047	3002913
 (Gly)VanA-B:AF192329:28835-29863:1029	3000013
 (Gly)VanA-D:AY082011:5901-6932:1032	3000005
-(Gly)VanA-G:AY271782:157-606:450	3007029
+(Gly)VanA-G:AY271782:157-606:450	3000010
 (Gly)VanA-M:FJ349556:4857-5888:1032	3002911
 (Gly)VanAE-Pp:AF155139:5940-6971:1032	3002908
 (Gly)VanB:AY655710:1-1029:1029	3000013
@@ -1632,7 +1632,7 @@ Original ID	ARO
 (MLS)ErmZ:AM709783:2817-3665:849	3004605
 (MLS)LinA:M14039:413-898:486	3002835
 (MLS)LinB:AJ238249:127-930:804	3002836
-(MLS)LmrA:X59926:318-1763:1446	3003955
+(MLS)LmrA:X59926:318-1763:1446	3003028
 (MLS)LmrB:X62867:361-1197:837	3001305
 (MLS)LmrB:X79146:27840-28679:840	3001305
 (MLS)LnuB:KC688833:1-804:804	3002836
@@ -1647,7 +1647,7 @@ Original ID	ARO
 (MLS)MphA:DQ445270:1626-2531:1230	3000316
 (MLS)MphB:D85892:1159-2067:909	3000318
 (MLS)MphC:AF167161:5665-6564:900	3000319
-(MLS)MphD:NC_017312:2291580-2292413:834	3004581
+(MLS)MphD:NC_017312:2291580-2292413:834	3000333
 (MLS)MphE:JF769133:8777-9661:885	3003741
 (MLS)MsrA:AY591760:274-1740:1467	3000251
 (MLS)MsrC:AY004350:496-1974:1479	3002819
@@ -1717,7 +1717,7 @@ Original ID	ARO
 (Rif)Arr5:EF660563:393-845:453	3002850
 (Rif)Arr7:FN397623:1189-1641:453	3002852
 (Rif)IRI:U56415:280-1719:1440	3002884
-(Rif)Rif:EF541029:530-2170:1641	3004047
+(Rif)Rif:EF541029:530-2170:1641	3004040
 (Sul)SulI:AF071413:6700-7539:840	3000410
 (Sul)SulII:EU360945:1617-2432:816	3000412
 (Sul)SulIII:HQ875016:7396-8187:792	3000413
@@ -1728,7 +1728,7 @@ Original ID	ARO
 (Tet)Tet-31:AJ250203:1651-2883:1233	3000476
 (Tet)Tet-32:DQ647324:181-2100:1920	3000196
 (Tet)Tet-33:AJ420072:22940-24163:1224	3000478
-(Tet)Tet-34:AB061440:306-770:465	3007840
+(Tet)Tet-34:AB061440:306-770:465	3002870
 (Tet)Tet-35:AF353562:2213-3322:1110	3000481
 (Tet)Tet-36:AJ514254:2534-4456:1923	3000197
 (Tet)Tet-37:AF540889:1-327:327	3002871
@@ -1758,7 +1758,7 @@ Original ID	ARO
 (Tet)TetR:HF545434:53576-54226:651	3003479
 (Tet)TetS:L09756:447-2372:1926	3000192
 (Tet)TetT:L42544:478-2433:1956	3000193
-(Tet)TetU:U01917:413-730:318	3002907
+(Tet)TetU:U01917:413-730:318	3004650
 (Tet)TetV:AF030344:462-1721:1260	3000181
 (Tet)TetW:AJ222769:3687-5606:1920	3000194
 (Tet)TetX:M37699:586-1752:1167	3000205
@@ -9373,7 +9373,7 @@ leuO	3003843
 lfrA	3003967
 lin	3004651
 linG	3002879
-lmr(A)_1_X59926	3003955
+lmr(A)_1_X59926	3003028
 lmr(B)_1_X62867	3001305
 lmrB	3002813
 lmrC	3002881
@@ -9866,7 +9866,7 @@ tet(32)_2_EF626943	3000196
 tet(33)	3000478
 tet(33)_1_AY255627	3000478
 tet(33)_2_DQ390458	3000478
-tet(34)_1_AB061440	3007840
+tet(34)_1_AB061440	3002870
 tet(35)	3000481
 tet(35)_1_AF353562	3000481
 tet(36)	3000197

argnorm/data/manual_curation/argannot_curation.tsv

@@ -1,4 +1,14 @@
-Original ID	ARO	Gene Name in CARD	Description
-(Phe)cpt_strepv:U09991:AAB36569:1412-1948:537	3000249	chloramphenicol phosphotransferase	Parent ARO mapping
-(Tet)tetH:EF460464:6286-7839:1554	3000175	tet(H)	Loose RGI mapping. Mapped incorrectly to ARO:3004797.
-(AGly)aadC:V01282:225-701:477	3000225	ANT(6)	Parent ARO mapping
+Original ID	ARO	Gene Name in CARD	Description
+(Phe)cpt_strepv:U09991:AAB36569:1412-1948:537	3000249	chloramphenicol phosphotransferase	Parent ARO mapping
+(AGly)aadC:V01282:225-701:477	3000225	ANT(6)	Parent ARO mapping
+(Gly)vanA-G:AY271782:157-606:450	3000010		
+(MLS)lmr(A):X59926:318-1763:1446	3003028		
+(MLS)mph(D):NC_017312:2291580-2292413:834	3000333		
+(Rif)rif:EF541029:530-2170:1641	3004040		
+(Tet)tet(34):AB061440:306-770:465	3002870		
+(Tet)tetH:EF460464:6286-7839:1554	3000175		
+(Tet)tetU:U01917:413-730:318	3004650		
+(TetracenomycinC)tcmA:NG_048121:101-1717:1617	3003554		
+(Phe)cpt:NG_051909:101-631:531	3000249		
+(Fcyn)FomC:AB016934:10868-11656:789	3004246		
+(Flq)crpP:NG_062203:WP_033179079:101-298:198	3004467		

argnorm/data/manual_curation/resfinderfg_curation.tsv

@@ -1,4 +1,9 @@
-Original ID	ARO	Gene Name in CARD	Description
-UDP-N-acetylmuramoyl-tripeptide--D-alanyl-D-alanine ligase|KF629588.1|pediatric_fecal_sample|CYC	3003970	D-Ala-D-Ala ligase	
-UDP-N-acetylmuramoyl-tripeptide--D-alanyl-D-alanine ligase|KF629153.1|pediatric_fecal_sample|CYC	3003970	D-Ala-D-Ala ligase
-UDP-N-acetylmuramoyl-tripeptide--D-alanyl-D-alanine ligase|KJ695568.1|Agricultural soil|CYC	3003970	D-Ala-D-Ala ligase
+Original ID	ARO	Gene Name in CARD	Description
+UDP-N-acetylmuramoyl-tripeptide--D-alanyl-D-alanine ligase|KF629588.1|pediatric_fecal_sample|CYC	3003970	D-Ala-D-Ala ligase	
+UDP-N-acetylmuramoyl-tripeptide--D-alanyl-D-alanine ligase|KF629153.1|pediatric_fecal_sample|CYC	3003970	D-Ala-D-Ala ligase
+UDP-N-acetylmuramoyl-tripeptide--D-alanyl-D-alanine ligase|KJ695568.1|Agricultural soil|CYC	3003970	D-Ala-D-Ala ligase
+UDP-N-acetylmuramoyl-tripeptide--D-alanyl-D-alanine ligase|KX125757.1|human_gut|CYC	3003970
+UDP-N-acetylmuramoyl-tripeptide--D-alanyl-D-alanine ligase|KX125843.1|human_gut|CYC	3003970
+UDP-N-acetylmuramoyl-tripeptide--D-alanyl-D-alanine ligase|KF627869.1|pediatric_fecal_sample|CYC	3003970
+UDP-N-acetylmuramoyl-tripeptide--D-alanyl-D-alanine ligase|KF629229.1|pediatric_fecal_sample|CYC	3003970
+UDP-N-acetylmuramoyl-tripeptide--D-alanyl-D-alanine ligase|KF630034.1|pediatric_fecal_sample|CYC	3003970

argnorm/data/manual_curation/sarg_curation.tsv

@@ -1,4 +1,5 @@
-Original ID	ARO	Gene Name in CARD	Description
-gb|AAG57600.1|ARO:3000318|mphB	3000318	mphB	
-AM180355.1.gene2260.p01	3000250	ErmC	
-gb|AUW34359.1|ARO:3004445|RSA-2	3005440	RSA2 beta-lactamase	ARO number of RSA2 had been changed.
+Original ID	ARO	Gene Name in CARD	Description
+gb|AAG57600.1|ARO:3000318|mphB	3000318	mphB	
+AM180355.1.gene2260.p01	3000250	ErmC	
+gb|AUW34359.1|ARO:3004445|RSA-2	3005440	RSA2 beta-lactamase	ARO number of RSA2 had been changed.
+gb|AAB08924.1|ARO:3004650|tetU	3004650		

db_harmonisation/README.md

@@ -179,6 +179,10 @@ cd .. && rm -r tmp
 
 Genes from ARG annotation outputs are mapped to ARO accessions using ARO annotation tables. ARO annotation tables are constructed using the RGI (Alcock et al., 2023) and manual curation. All databases except MEGARes v3.0 are handled as amino acid files, where all ARG sequences (coding sequences) in the databases are translated to amino acid form using BioPython (Cock et al., 2009). The amino acid files are processed by RGI using the ‘protein’ mode to map genes to the ARO. The ‘Original ID’ (gene name in ARG annotation database), ‘Best_Hit_ARO’ (gene name in CARD) and ‘ARO’ (ARO accession) columns from the RGI output are specifically chosen from the RGI output, with the ‘Best_Hit_ARO’ column renamed to ‘Gene Name in CARD’ , to form the automated annotation tables. Genes which were not given an ARO mapping by RGI are manually assigned an ARO accession. The manual curation and automated annotation tables are combined to produce ARO annotation tables.
 
+All RGI hits are checked for mapping accuracy using `check_mapping_accuracy.py`. This is done by comparing the drug categorization of the ARO assigned to the ARG with the drug categorization assigned to the ARG by its original database. A list of ARGs with mismatched mappings are included in `manual_curation/`. Mismatched hits are manually curated to correct ARGs and manual curation files are updated. Some mismatched hits are marked `correct` or `incorrect`. `correct` hits have a drug category mismatch but have an ARO mapping that's determined to be correct. `incorrect` hits have a drug category mismatch and a manually curated ARO can't be found for them. Both `correct` and `incorrect` hits should not be included in the final manual curation. For megares, a file called `megares_meta_biocide_and_virulence_genes.tsv` is also created with metal, biocide, and virulence genes. These should not be mapped to any ARO and can be manually curated to no ARO. Metal, biocide, and virulence genes are directly added to manual curation files for other databases. Groot is derived from resfinder and argannot. All loose hit manual curation for resfinder and argannot should also be manually updated for groot.
+
+A list of all mismatched hits, with their manual curation can be found in `all_mismatched_hits.tsv`
+
 ## Handling ResFinder
 The ResFinder v4.0 is a notable example as it contains forty instances of gene clusters or sequences with multiple coding sequences within (e.g. vanM_1_FJ349556 contains seven coding sequences). As RGI maps a single gene from the ResFinder database to multiple AROs (corresponding to different coding sequences within) in cases with gene clusters, gene clusters are manually assigned ARO accessions corresponding to specific gene clusters. Reverse complement sequences within ResFinder are also manually assigned ARO accessions.