{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2026,5,23]],"date-time":"2026-05-23T19:05:06Z","timestamp":1779563106735,"version":"3.53.1"},"reference-count":31,"publisher":"Oxford University Press (OUP)","issue":"5","license":[{"start":{"date-parts":[[2026,5,9]],"date-time":"2026-05-09T00:00:00Z","timestamp":1778284800000},"content-version":"vor","delay-in-days":8,"URL":"https:\/\/creativecommons.org\/licenses\/by\/4.0\/"}],"content-domain":{"domain":[],"crossmark-restriction":false},"short-container-title":[],"published-print":{"date-parts":[[2026,5,3]]},"abstract":"<jats:title>Abstract<\/jats:title>\n                  <jats:sec>\n                    <jats:title>Motivation<\/jats:title>\n                    <jats:p>Accurate prediction of HIV drug resistance from viral sequences is critical for optimizing antiretroviral therapy. Traditional machine-learning approaches using binary mutation encoding achieve strong accuracy but may fail to capture epistatic interactions and structural features relevant to resistance mechanisms. Protein language models (PLMs) offer learned representations encoding evolutionary and structural information, but have not been systematically benchmarked for HIV resistance prediction across the modern antiretroviral drug set.<\/jats:p>\n                  <\/jats:sec>\n                  <jats:sec>\n                    <jats:title>Results<\/jats:title>\n                    <jats:p>We evaluated ESM-2 (650\u2009M parameters) with attention-weighted pooling for predicting resistance to 18 drugs across three classes (protease inhibitors, NRTIs, NNRTIs) on the Stanford HIVDB dataset (n\u2009=\u20096308 sequences). Attention-weighted ESM-2 embeddings significantly outperformed XGBoost baselines with binary mutation encoding (mean AUC 0.968 versus 0.955, P\u2009=\u2009.0017), with gains across 15 of 18 drugs and the largest improvements for drugs with complex resistance patterns. Attention weights showed 2.48-fold enrichment at known drug-resistance-mutation positions (P\u2009&amp;lt;\u2009.05 for 63% of drugs; NRTIs strongest at 4.20-fold). External validation on a 20% holdout showed minimal degradation (AUC 0.934). Benchmarking against ESM C 600M and ESM-1v confirmed performance is robust to PLM choice (mean AUC 0.942\u20130.946 across backbones). Performance was maintained across HIV-1 subtypes (B 0.924; B-divergent 0.900; non-B 0.884) and a temporal holdout (AUC 0.930).<\/jats:p>\n                  <\/jats:sec>\n                  <jats:sec>\n                    <jats:title>Availability and implementation<\/jats:title>\n                    <jats:p>Source code is available at https:\/\/github.com\/hayden-farquhar\/HIV-ESM-2 under an MIT license and archived at https:\/\/doi.org\/10.5281\/zenodo.19466629. Stanford HIVDB genotype\u2013phenotype data are publicly available at https:\/\/hivdb.stanford.edu\/.<\/jats:p>\n                  <\/jats:sec>","DOI":"10.1093\/bioinformatics\/btag260","type":"journal-article","created":{"date-parts":[[2026,5,7]],"date-time":"2026-05-07T11:36:59Z","timestamp":1778153819000},"source":"Crossref","is-referenced-by-count":0,"title":["Protein language model embeddings improve HIV drug resistance prediction: a comprehensive benchmark with attention-based interpretability"],"prefix":"10.1093","volume":"42","author":[{"ORCID":"https:\/\/orcid.org\/0009-0002-6226-440X","authenticated-orcid":false,"given":"Hayden","family":"Farquhar","sequence":"first","affiliation":[{"name":"Independent Researcher , Finley, New South Wales,","place":["Australia"]}],"role":[{"vocabulary":"crossref","role":"author"}]}],"member":"286","published-online":{"date-parts":[[2026,5,9]]},"reference":[{"key":"2026052314563736900_btag260-B1","doi-asserted-by":"crossref","first-page":"3850","DOI":"10.1093\/nar\/gkg575","article-title":"Geno2pheno: estimating phenotypic drug resistance from HIV-1 genotypes","volume":"31","author":"Beerenwinkel","year":"2003","journal-title":"Nucleic Acids Res"},{"key":"2026052314563736900_btag260-B2","doi-asserted-by":"crossref","first-page":"42","DOI":"10.1038\/s41572-023-00452-3","article-title":"HIV infection","volume":"9","author":"Bekker","year":"2023","journal-title":"Nat Rev Dis Primers"},{"key":"2026052314563736900_btag260-B3","doi-asserted-by":"crossref","first-page":"e1008873","DOI":"10.1371\/journal.pcbi.1008873","article-title":"Using machine learning and big data to explore the drug resistance landscape in HIV","volume":"17","author":"Blassel","year":"2021","journal-title":"PLoS Comput Biol"},{"key":"2026052314563736900_btag260-B4","doi-asserted-by":"crossref","first-page":"56","DOI":"10.1016\/j.coviro.2021.09.009","article-title":"Drug resistance mutations in HIV: new bioinformatics approaches and challenges","volume":"51","author":"Blassel","year":"2021","journal-title":"Curr Opin Virol"},{"key":"2026052314563736900_btag260-B5","first-page":"785","author":"Chen","year":"2016"},{"key":"2026052314563736900_btag260-B6","doi-asserted-by":"crossref","first-page":"347","DOI":"10.3390\/v3040347","article-title":"Clinical management of HIV drug resistance","volume":"3","author":"Cortez","year":"2011","journal-title":"Viruses"},{"key":"2026052314563736900_btag260-B7","first-page":"231","volume-title":"Human Immunodeficiency Virus Reverse Transcriptase","author":"Das","year":"2012"},{"key":"2026052314563736900_btag260-B8","doi-asserted-by":"crossref","first-page":"7112","DOI":"10.1109\/TPAMI.2021.3095381","article-title":"ProtTrans: toward understanding the language of life through self-supervised learning","volume":"44","author":"Elnaggar","year":"2022","journal-title":"IEEE Trans Pattern Anal Mach Intell"},{"key":"2026052314563736900_btag260-B9","author":"EvolutionaryScale","year":"2024"},{"key":"2026052314563736900_btag260-B10","doi-asserted-by":"crossref","first-page":"4310","DOI":"10.1021\/acs.jcim.3c02061","article-title":"Predicting antimicrobial peptides using ESMFold-predicted structures and ESM-2-based amino acid features with graph deep learning","volume":"64","author":"Garc\u00eda-Jacas","year":"2024","journal-title":"J Chem Inf Model"},{"key":"2026052314563736900_btag260-B11","doi-asserted-by":"publisher","first-page":"850","DOI":"10.1126\/science.ads0018","article-title":"Simulating 500 million years of evolution with a language model","volume":"387","author":"Hayes","year":"2025","journal-title":"Science"},{"key":"2026052314563736900_btag260-B12","doi-asserted-by":"publisher","first-page":"37456","DOI":"10.1038\/s41598-025-21223-y","article-title":"Transfer learning on protein language models improves antimicrobial peptide classification","volume":"15","author":"Georgoulis","year":"2025","journal-title":"Sci Rep"},{"key":"2026052314563736900_btag260-B13","doi-asserted-by":"crossref","first-page":"1123","DOI":"10.1126\/science.ade2574","article-title":"Evolutionary-scale prediction of atomic-level protein structure with a language model","volume":"379","author":"Lin","year":"2023","journal-title":"Science"},{"key":"2026052314563736900_btag260-B14","doi-asserted-by":"crossref","first-page":"1608","DOI":"10.1086\/503914","article-title":"Web resources for HIV type 1 genotypic-resistance test interpretation","volume":"42","author":"Liu","year":"2006","journal-title":"Clin Infect Dis"},{"key":"2026052314563736900_btag260-B15","first-page":"29287","article-title":"Language models enable zero-shot prediction of the effects of mutations on protein function","volume":"34","author":"Meier","year":"2021","journal-title":"Adv Neural Inf Process Syst"},{"key":"2026052314563736900_btag260-B16","first-page":"3449","article-title":"Drug resistance conferred by mutations outside the active site through alterations in the dynamic and structural ensemble of HIV-1 protease","volume":"144","author":"Paulsen","year":"2022","journal-title":"J Am Chem Soc"},{"key":"2026052314563736900_btag260-B17","doi-asserted-by":"crossref","first-page":"362","DOI":"10.1186\/s12859-018-2331-y","article-title":"Analysis of drug resistance in HIV protease","volume":"19","author":"Pawar","year":"2018","journal-title":"BMC Bioinformatics"},{"key":"2026052314563736900_btag260-B18","doi-asserted-by":"crossref","first-page":"920","DOI":"10.1128\/AAC.44.4.920-928.2000","article-title":"A novel phenotypic drug susceptibility assay for human immunodeficiency virus type 1","volume":"44","author":"Petropoulos","year":"2000","journal-title":"Antimicrob Agents Chemother"},{"key":"2026052314563736900_btag260-B19","doi-asserted-by":"crossref","first-page":"52","DOI":"10.1038\/nrg1246","article-title":"The causes and consequences of HIV evolution","volume":"5","author":"Rambaut","year":"2004","journal-title":"Nat Rev Genet"},{"key":"2026052314563736900_btag260-B20","doi-asserted-by":"crossref","first-page":"17355","DOI":"10.1073\/pnas.0607274103","article-title":"Genotypic predictors of human immunodeficiency virus type 1 drug resistance","volume":"103","author":"Rhee","year":"2006","journal-title":"Proc Natl Acad Sci U S A"},{"key":"2026052314563736900_btag260-B21","doi-asserted-by":"crossref","first-page":"693","DOI":"10.1016\/j.jmb.2008.10.071","article-title":"Structure and function of HIV-1 reverse transcriptase: molecular mechanisms of polymerization and inhibition","volume":"385","author":"Sarafianos","year":"2009","journal-title":"J Mol Biol"},{"key":"2026052314563736900_btag260-B22","first-page":"2882","article-title":"HIV-1 protease: structural perspectives on drug resistance","volume":"3","author":"Schiffer","year":"2011","journal-title":"Viruses"},{"key":"2026052314563736900_btag260-B23","doi-asserted-by":"crossref","first-page":"S51","DOI":"10.1086\/505356","article-title":"Rationale and uses of a public HIV drug-resistance database","volume":"194","author":"Shafer","year":"2006","journal-title":"J Infect Dis"},{"key":"2026052314563736900_btag260-B24","doi-asserted-by":"crossref","first-page":"497","DOI":"10.1186\/s12859-020-03825-7","article-title":"Evolution of drug resistance in HIV protease","volume":"21","author":"Shah","year":"2020","journal-title":"BMC Bioinformatics"},{"key":"2026052314563736900_btag260-B25","doi-asserted-by":"publisher","author":"Xu J, Shi Y, Lang L et al.","DOI":"10.48550\/arXiv.2510.03370,"},{"key":"2026052314563736900_btag260-B26","doi-asserted-by":"crossref","first-page":"560","DOI":"10.3390\/v12050560","article-title":"Drug resistance prediction using deep learning techniques on HIV-1 sequence data","volume":"12","author":"Steiner","year":"2020","journal-title":"Viruses"},{"key":"2026052314563736900_btag260-B27","author":"UNAIDS","year":"2025"},{"key":"2026052314563736900_btag260-B28","doi-asserted-by":"crossref","first-page":"1110","DOI":"10.3390\/v1031110","article-title":"HIV-1 protease: structural perspectives on drug resistance","volume":"1","author":"Weber","year":"2009","journal-title":"Viruses"},{"key":"2026052314563736900_btag260-B29","first-page":"559","article-title":"2022 update of the drug resistance mutations in HIV-1","volume":"30","author":"Wensing","year":"2022","journal-title":"Top Antivir Med"},{"key":"2026052314563736900_btag260-B30","author":"WHO","year":"2025"},{"key":"2026052314563736900_btag260-B31","doi-asserted-by":"crossref","first-page":"btad690","DOI":"10.1093\/bioinformatics\/btad690","article-title":"PLM-ARG: antibiotic resistance gene identification using a pretrained protein language model","volume":"39","author":"Wu","year":"2023","journal-title":"Bioinformatics"}],"container-title":["Bioinformatics"],"original-title":[],"language":"en","link":[{"URL":"https:\/\/academic.oup.com\/bioinformatics\/advance-article-pdf\/doi\/10.1093\/bioinformatics\/btag260\/68261076\/btag260.pdf","content-type":"application\/pdf","content-version":"am","intended-application":"syndication"},{"URL":"https:\/\/academic.oup.com\/bioinformatics\/article-pdf\/42\/5\/btag260\/68261076\/btag260.pdf","content-type":"application\/pdf","content-version":"vor","intended-application":"syndication"},{"URL":"https:\/\/academic.oup.com\/bioinformatics\/article-pdf\/42\/5\/btag260\/68261076\/btag260.pdf","content-type":"unspecified","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2026,5,23]],"date-time":"2026-05-23T18:56:46Z","timestamp":1779562606000},"score":1,"resource":{"primary":{"URL":"https:\/\/academic.oup.com\/bioinformatics\/article\/doi\/10.1093\/bioinformatics\/btag260\/8675121"}},"subtitle":[],"editor":[{"given":"Daisuke","family":"Kihara","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"editor"}]}],"short-title":[],"issued":{"date-parts":[[2026,5]]},"references-count":31,"journal-issue":{"issue":"5","published-print":{"date-parts":[[2026,5,3]]}},"URL":"https:\/\/doi.org\/10.1093\/bioinformatics\/btag260","relation":{},"ISSN":["1367-4803","1367-4811"],"issn-type":[{"value":"1367-4803","type":"print"},{"value":"1367-4811","type":"electronic"}],"subject":[],"published-other":{"date-parts":[[2026,5]]},"published":{"date-parts":[[2026,5]]},"article-number":"btag260"}}