{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2026,8,18]],"date-time":"2026-08-18T04:10:00Z","timestamp":1787026200827,"version":"build-2736575974"},"reference-count":53,"publisher":"Frontiers Media SA","license":[{"start":{"date-parts":[[2023,8,17]],"date-time":"2023-08-17T00:00:00Z","timestamp":1692230400000},"content-version":"vor","delay-in-days":0,"URL":"https:\/\/creativecommons.org\/licenses\/by\/4.0\/"}],"content-domain":{"domain":["frontiersin.org"],"crossmark-restriction":true},"short-container-title":["Front. Bioinform."],"abstract":"<jats:p>Major histocompatibility complex Class I (MHC-I) molecules bind to peptides derived from intracellular antigens and present them on the surface of cells, allowing the immune system (T cells) to detect them. Elucidating the process of this presentation is essential for regulation and potential manipulation of the cellular immune system. Predicting whether a given peptide binds to an MHC molecule is an important step in the above process and has motivated the introduction of many computational approaches to address this problem. NetMHCPan, a pan-specific model for predicting binding of peptides to any MHC molecule, is one of the most widely used methods which focuses on solving this binary classification problem using shallow neural networks. The recent successful results of Deep Learning (DL) methods, especially Natural Language Processing (NLP-based) pretrained models in various applications, including protein structure determination, motivated us to explore their use in this problem. Specifically, we consider the application of deep learning models pretrained on large datasets of protein sequences to predict MHC Class I-peptide binding. Using the standard performance metrics in this area, and the same training and test sets, we show that our models outperform NetMHCpan4.1, currently considered as the-state-of-the-art.<\/jats:p>","DOI":"10.3389\/fbinf.2023.1207380","type":"journal-article","created":{"date-parts":[[2023,8,18]],"date-time":"2023-08-18T01:19:39Z","timestamp":1692321579000},"update-policy":"https:\/\/doi.org\/10.3389\/crossmark-policy","source":"Crossref","is-referenced-by-count":18,"title":["Improved prediction of MHC-peptide binding using protein language models"],"prefix":"10.3389","volume":"3","author":[{"given":"Nasser","family":"Hashemi","sequence":"first","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Boran","family":"Hao","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Mikhail","family":"Ignatov","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Ioannis Ch.","family":"Paschalidis","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Pirooz","family":"Vakili","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Sandor","family":"Vajda","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Dima","family":"Kozakov","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]}],"member":"1965","published-online":{"date-parts":[[2023,8,17]]},"reference":[{"key":"B1","doi-asserted-by":"publisher","first-page":"315","DOI":"10.1016\/j.immuni.2017.02.007","article-title":"Mass spectrometry profiling of hla-associated peptidomes in mono-allelic cells enables more accurate epitope prediction","volume":"46","author":"Abelin","year":"2017","journal-title":"Immunity"},{"key":"B2","article-title":"Publicly available clinical bert embeddings","author":"Alsentzer","year":"2019"},{"key":"B3","doi-asserted-by":"publisher","first-page":"2459","DOI":"10.1074\/mcp.tir119.001658","article-title":"Nnalign_ma; mhc peptidome deconvolution for accurate mhc binding motif characterization and improved t-cell epitope predictions","volume":"18","author":"Alvarez","year":"2019","journal-title":"Mol. Cell. Proteomics"},{"key":"B4","doi-asserted-by":"publisher","first-page":"1962","DOI":"10.4049\/jimmunol.1900918","article-title":"Combining three-dimensional modeling with artificial intelligence to increase specificity and precision in peptide\u2013mhc binding predictions","volume":"205","author":"Aranha","year":"2020","journal-title":"J. Immunol."},{"key":"B5","doi-asserted-by":"publisher","first-page":"e1005725","DOI":"10.1371\/journal.pcbi.1005725","article-title":"Deciphering hla-i motifs across hla peptidomes improves neo-antigen predictions and identifies allostery regulating hla specificity","volume":"13","author":"Bassani-Sternberg","year":"2017","journal-title":"PLoS Comput. Biol."},{"key":"B6","doi-asserted-by":"publisher","first-page":"2492","DOI":"10.4049\/jimmunol.1600808","article-title":"Unsupervised hla peptidome deconvolution improves ligand prediction accuracy and predicts cooperative effects in peptide\u2013hla interactions","volume":"197","author":"Bassani-Sternberg","year":"2016","journal-title":"J. Immunol."},{"key":"B7","doi-asserted-by":"publisher","first-page":"7","DOI":"10.1186\/s12859-018-2561-z","article-title":"Predicting peptide presentation by major histocompatibility complex class i: an improved machine learning approach to the immunopeptidome","volume":"20","author":"Boehm","year":"2019","journal-title":"BMC Bioinforma."},{"key":"B8","doi-asserted-by":"crossref","DOI":"10.18653\/v1\/D15-1075","article-title":"A large annotated corpus for learning natural language inference","author":"Bowman","year":"2015"},{"key":"B9","doi-asserted-by":"publisher","first-page":"304","DOI":"10.1007\/s00251-005-0798-y","article-title":"Automated generation and evaluation of specific mhc binding predictive tools: arb matrix applications","volume":"57","author":"Bui","year":"2005","journal-title":"Immunogenetics"},{"key":"B10","doi-asserted-by":"publisher","first-page":"3105","DOI":"10.1074\/mcp.o115.052431","article-title":"Analysis of major histocompatibility complex (mhc) immunopeptidomes using mass spectrometry","volume":"14","author":"Caron","year":"2015","journal-title":"Mol. Cell. Proteomics"},{"key":"B11","article-title":"Bertmhc: improves mhc-peptide class ii interaction prediction with transformer and multiple instance learning","author":"Cheng","year":"2020","journal-title":"bioRxiv"},{"key":"B12","doi-asserted-by":"crossref","first-page":"8604","DOI":"10.1109\/ICASSP.2013.6639345","article-title":"Recent advances in deep learning for speech research at microsoft","volume-title":"2013 IEEE international conference on acoustics, speech and signal processing","author":"Deng","year":"2013"},{"key":"B13","article-title":"Bert: pre-training of deep bidirectional transformers for language understanding","author":"Devlin","year":"2018"},{"key":"B14","doi-asserted-by":"publisher","first-page":"3572","DOI":"10.1021\/jm010021j","article-title":"Toward the quantitative prediction of t-cell epitopes: comfa and comsia studies of peptides with affinity for the class i mhc molecule hla-a* 0201","volume":"44","author":"Doytchinova","year":"2001","journal-title":"J. Med. Chem."},{"key":"B15","doi-asserted-by":"publisher","first-page":"1922","DOI":"10.1002\/prot.26209","article-title":"Assessing the binding properties of casp14 targets and models","volume":"89","author":"Egbert","year":"2021","journal-title":"Proteins Struct. Funct. Bioinforma."},{"key":"B16","article-title":"Prottrans: towards cracking the language of life\u2019s code through self-supervised deep learning and high performance computing","author":"Elnaggar","year":"2020"},{"key":"B17","doi-asserted-by":"publisher","first-page":"113278","DOI":"10.1016\/j.yexcr.2022.113278","article-title":"Application of deep learning methods: from molecular modelling to patient classification","volume":"418","author":"Fu","year":"2022","journal-title":"Exp. Cell. Res."},{"key":"B18","article-title":"Interpreting bert architecture predictions for peptide presentation by mhc class i proteins","author":"Gasser","year":"2021"},{"key":"B19","article-title":"Improved docking of protein models by a combination of alphafold2 and cluspro","author":"Ghani","year":"2021","journal-title":"bioRxiv"},{"key":"B20","doi-asserted-by":"crossref","DOI":"10.20537\/2076-7633-2020-12-6-1383-1395","article-title":"Application of an ensemble of neural networks and methods of statistical mechanics to predict binding of a peptide to a major histocompatibility complex","author":"Grebenkin","year":"2020","journal-title":"Comput. Res. Model"},{"key":"B21","first-page":"5469","article-title":"Conflibert: a pre-trained language model for political conflict and violence","volume-title":"Proceedings of the 2022 conference of the north American chapter of the association for computational linguistics: human language technologies","author":"Hu","year":"2022"},{"key":"B22","doi-asserted-by":"crossref","first-page":"31","DOI":"10.1142\/9789811258589_0002","article-title":"Application of sequence embedding in protein sequence-based predictions","volume-title":"Machine learning in bioinformatics of protein sequences: algorithms, databases and resources for modern protein bioinformatics","author":"Ibtehaz","year":"2023"},{"key":"B23","doi-asserted-by":"publisher","first-page":"3198","DOI":"10.1016\/j.csbj.2021.05.039","article-title":"Representation learning applications in biological sequence analysis","volume":"19","author":"Iuchi","year":"2021","journal-title":"Comput. Struct. Biotechnol. J."},{"key":"B24","volume-title":"Immunobiology","author":"Janeway","year":"2001"},{"key":"B25","doi-asserted-by":"publisher","first-page":"583","DOI":"10.1038\/s41586-021-03819-2","article-title":"Highly accurate protein structure prediction with alphafold","volume":"596","author":"Jumper","year":"2021","journal-title":"Nature"},{"key":"B26","first-page":"24","article-title":"High accuracy protein structure prediction using deep learning","volume":"22","author":"Jumper","year":"2020","journal-title":"Fourteenth Crit. Assess. Tech. Protein Struct. Predict."},{"key":"B27","doi-asserted-by":"publisher","first-page":"241","DOI":"10.1016\/j.ymssp.2017.11.024","article-title":"A review on the application of deep learning in system health management","volume":"107","author":"Khan","year":"2018","journal-title":"Mech. Syst. Signal Process."},{"key":"B28","doi-asserted-by":"publisher","first-page":"436","DOI":"10.1038\/nature14539","article-title":"Deep learning","volume":"521","author":"LeCun","year":"2015","journal-title":"nature"},{"key":"B29","doi-asserted-by":"publisher","first-page":"1800","DOI":"10.1002\/prot.26222","article-title":"Prediction of protein assemblies, the next frontier: the casp14-capri experiment","volume":"89","author":"Lensink","year":"2021","journal-title":"Proteins Struct. Funct. Bioinforma."},{"key":"B30","article-title":"Evolutionary-scale prediction of atomic level protein structure with a language model","author":"Lin","year":"2022","journal-title":"bioRxiv"},{"key":"B31","doi-asserted-by":"publisher","first-page":"1","DOI":"10.1016\/j.csbj.2018.11.004","article-title":"Fates of cd8+ t cells in tumor microenvironment","volume":"17","author":"Maimela","year":"2019","journal-title":"Comput. Struct. Biotechnol. J."},{"key":"B32","first-page":"570","article-title":"A framework for multiple-instance learning","author":"Maron","year":"1998","journal-title":"Adv. neural Inf. Process. Syst."},{"key":"B33","doi-asserted-by":"publisher","first-page":"101566","DOI":"10.1016\/j.eml.2021.101566","article-title":"Predicting mechanically driven full-field quantities of interest with deep learning-based metamodels","volume":"50","author":"Mohammadzadeh","year":"2021","journal-title":"Extreme Mech. Lett."},{"key":"B34","doi-asserted-by":"publisher","first-page":"1007","DOI":"10.1110\/ps.0239403","article-title":"Reliable prediction of t-cell epitopes using neural networks with novel sequence representations","volume":"12","author":"Nielsen","year":"2003","journal-title":"Protein Sci."},{"key":"B35","doi-asserted-by":"publisher","first-page":"42","DOI":"10.1016\/j.cels.2020.06.010","article-title":"Mhcflurry 2.0: improved pan-allele prediction of mhc class i-presented peptides by incorporating antigen processing","volume":"11","author":"O\u2019Donnell","year":"2020","journal-title":"Cell. Syst."},{"key":"B36","doi-asserted-by":"publisher","first-page":"518","DOI":"10.1016\/j.csbj.2020.12.039","article-title":"Computational design of sars-cov-2 spike glycoproteins to increase immunogenicity by t cell epitope engineering","volume":"19","author":"Ong","year":"2021","journal-title":"Comput. Struct. Biotechnol. J."},{"key":"B37","doi-asserted-by":"publisher","first-page":"163","DOI":"10.4049\/jimmunol.152.1.163","article-title":"Scheme for ranking potential hla-a2 binding peptides based on independent binding of individual peptide side-chains","volume":"152","author":"Parker","year":"1994","journal-title":"J. Immunol."},{"key":"B38","doi-asserted-by":"publisher","first-page":"9689","DOI":"10.1101\/676825","article-title":"Evaluating protein transfer learning with tape","volume":"32","author":"Rao","year":"2019","journal-title":"Adv. Neural Inf. Process. Syst."},{"key":"B39","doi-asserted-by":"crossref","DOI":"10.1101\/2021.02.12.430858","article-title":"Msa transformer","author":"Rao","year":"2021"},{"key":"B40","doi-asserted-by":"publisher","DOI":"10.1101\/2020.12.15.422761","article-title":"Transformer protein language models are unsupervised structure learners","author":"Rao","year":"2020","journal-title":"bioRxiv"},{"key":"B41","doi-asserted-by":"publisher","first-page":"W449","DOI":"10.1093\/nar\/gkaa379","article-title":"Netmhcpan-4.1 and netmhciipan-4.0: improved predictions of mhc antigen presentation by concurrent motif deconvolution and integration of ms mhc eluted ligand data","volume":"48","author":"Reynisson","year":"2020","journal-title":"Nucleic acids Res."},{"key":"B42","first-page":"622803","article-title":"Biological structure and function emerge from scaling unsupervised learning to 250 million protein sequences","author":"Rives","year":"2019"},{"key":"B43","doi-asserted-by":"publisher","first-page":"D1237","DOI":"10.1093\/nar\/gkx664","article-title":"The systemhc atlas project","volume":"46","author":"Shao","year":"2018","journal-title":"Nucleic acids Res."},{"key":"B44","doi-asserted-by":"publisher","first-page":"2000","DOI":"10.1016\/j.csbj.2020.07.008","article-title":"Methods for sequence and structural analysis of b and t cell receptor repertoires","volume":"18","author":"Teraguchi","year":"2020","journal-title":"Comput. Struct. Biotechnol. J."},{"key":"B45","doi-asserted-by":"crossref","DOI":"10.1007\/978-1-4419-9863-7_942","article-title":"Blocks substitution matrix (blosum)","volume-title":"Encyclopedia of systems biology","author":"Tong","year":"2013"},{"key":"B46","doi-asserted-by":"publisher","first-page":"285","DOI":"10.1016\/0092-8674(90)90366-m","article-title":"Assembly of mhc class i molecules analyzed in vitro","volume":"62","author":"Townsend","year":"1990","journal-title":"Cell."},{"key":"B47","article-title":"Attention is all you need","author":"Vaswani","year":"2017"},{"key":"B48","article-title":"Graph attention networks","author":"Veli\u010dkovi\u0107","year":"2017"},{"key":"B49","doi-asserted-by":"publisher","first-page":"D339","DOI":"10.1093\/nar\/gky1006","article-title":"The immune epitope database (iedb): 2018 update","volume":"47","author":"Vita","year":"2019","journal-title":"Nucleic acids Res."},{"key":"B50","doi-asserted-by":"publisher","first-page":"1","DOI":"10.1155\/2018\/7068349","article-title":"Deep learning for computer vision: a brief review","volume":"2018","author":"Voulodimos","year":"2018","journal-title":"Comput. Intell. Neurosci."},{"key":"B51","doi-asserted-by":"publisher","first-page":"i278","DOI":"10.1093\/bioinformatics\/btz330","article-title":"Deepligand: accurate prediction of mhc class i ligands using peptide embedding","volume":"35","author":"Zeng","year":"2019","journal-title":"Bioinformatics"},{"key":"B52","doi-asserted-by":"publisher","first-page":"959","DOI":"10.1142\/s0219720006002314","article-title":"Optimally-connected hidden markov models for predicting mhc-binding peptides","volume":"4","author":"Zhang","year":"2006","journal-title":"J. Bioinforma. Comput. Biol."},{"key":"B53","doi-asserted-by":"publisher","first-page":"690049","DOI":"10.3389\/fgene.2021.690049","article-title":"Graph neural networks and their current applications in bioinformatics","volume":"12","author":"Zhang","year":"2021","journal-title":"Front. Genet."}],"container-title":["Frontiers in Bioinformatics"],"original-title":[],"link":[{"URL":"https:\/\/www.frontiersin.org\/articles\/10.3389\/fbinf.2023.1207380\/full","content-type":"unspecified","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2023,8,18]],"date-time":"2023-08-18T01:19:46Z","timestamp":1692321586000},"score":1,"resource":{"primary":{"URL":"https:\/\/www.frontiersin.org\/articles\/10.3389\/fbinf.2023.1207380\/full"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2023,8,17]]},"references-count":53,"alternative-id":["10.3389\/fbinf.2023.1207380"],"URL":"https:\/\/doi.org\/10.3389\/fbinf.2023.1207380","relation":{"has-preprint":[{"id-type":"doi","id":"10.1101\/2022.02.11.479844","asserted-by":"object"}]},"ISSN":["2673-7647"],"issn-type":[{"value":"2673-7647","type":"electronic"}],"subject":[],"published":{"date-parts":[[2023,8,17]]},"article-number":"1207380"}}