{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2026,10,24]],"date-time":"2026-10-24T00:23:28Z","timestamp":1792801408677,"version":"4.5.0"},"reference-count":38,"publisher":"Springer Science and Business Media LLC","issue":"1","license":[{"start":{"date-parts":[[2026,4,27]],"date-time":"2026-04-27T00:00:00Z","timestamp":1777248000000},"content-version":"tdm","delay-in-days":0,"URL":"https:\/\/creativecommons.org\/licenses\/by\/4.0"},{"start":{"date-parts":[[2026,7,24]],"date-time":"2026-07-24T00:00:00Z","timestamp":1784851200000},"content-version":"vor","delay-in-days":88,"URL":"https:\/\/creativecommons.org\/licenses\/by\/4.0"}],"funder":[{"name":"King Abdullah University of Science and Technology (KAUST), Thuwal, Saudi Arabia","id":[{"id":"https:\/\/ror.org\/01q3tbs38","id-type":"ROR","asserted-by":"crossref"}]},{"DOI":"10.13039\/501100004047","name":"Karolinska Institute","doi-asserted-by":"crossref","id":[{"id":"10.13039\/501100004047","id-type":"DOI","asserted-by":"crossref"}]}],"content-domain":{"domain":["link.springer.com"],"crossmark-restriction":false},"short-container-title":["J Cheminform"],"abstract":"<jats:title>Abstract<\/jats:title>\n                  <jats:sec>\n                    <jats:title>Introduction<\/jats:title>\n                    <jats:p>Predicting drug\u2013target interactions remains a significant challenge in drug development and lead optimization. Recent advances have leveraged machine learning algorithms to model drug\u2013target interactions from molecular and sequence data.<\/jats:p>\n                  <\/jats:sec>\n                  <jats:sec>\n                    <jats:title>Materials and Methods<\/jats:title>\n                    <jats:p>In this work, we use Evolutionary Scale Modeling (ESM-3) to construct a transformer-based protein language representation for drug\u2013target interaction prediction. We introduce LEP-AD (Language Embedding of Proteins and Attention to Drugs), a modular architecture that combines pretrained protein language models with graph-based molecular encoders to predict binding affinity values.<\/jats:p>\n                  <\/jats:sec>\n                  <jats:sec>\n                    <jats:title>Results<\/jats:title>\n                    <jats:p>We systematically benchmark LEP-AD alongside a range of established deep learning methods across multiple datasets\u2014Davis, KIBA, DTC, Metz, ToxCast, and STITCH. To assess predictive validity, we compare model-derived rankings of drug\u2013target interactions with experimental results reported in the literature. In addition, we perform new experimental assays to evaluate the binding of three ATP-competitive Src kinase inhibitors\u2014Dasatinib, UM-164, and Saracatinib\u2014where experimentally measured IC\u2085\u2080 and pK\u1d62 values are consistent with the predicted rankings.<\/jats:p>\n                  <\/jats:sec>\n                  <jats:sec>\n                    <jats:title>Conclusion<\/jats:title>\n                    <jats:p>In summary, our benchmark highlights the strengths and limitations of current drug\u2013target interaction models across diverse datasets and evaluation settings. The results emphasize the impact of pretrained protein and molecular representations on predictive performance and illustrate the persistent challenges of generalization, while the modular LEP-AD framework provides a flexible reference point for comparative evaluation.<\/jats:p>\n                  <\/jats:sec>\n                  <jats:sec>\n                    <jats:title>Scientific Contribution<\/jats:title>\n                    <jats:p>This study presents LEP-AD, a modular deep learning framework for drug\u2013target interaction prediction that integrates pretrained protein language representations with graph-based molecular encoders. Beyond introducing the architecture, we provide a systematic benchmark under similarity-aware evaluation settings and experimental validation, highlighting the impact of pretrained protein embeddings on predictive behavior across diverse datasets.<\/jats:p>\n                  <\/jats:sec>","DOI":"10.1186\/s13321-026-01167-9","type":"journal-article","created":{"date-parts":[[2026,4,27]],"date-time":"2026-04-27T19:46:08Z","timestamp":1777319168000},"update-policy":"https:\/\/doi.org\/10.1007\/springer_crossmark_policy","source":"Crossref","is-referenced-by-count":1,"title":["LEP-AD: language embedding of proteins and attention to drugs predicts drug-target interactions"],"prefix":"10.1186","volume":"18","author":[{"given":"Reem","family":"Alsulami","sequence":"first","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Robert","family":"Lehmann","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Anuj","family":"Daga","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Sumeer A.","family":"Khan","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Raik","family":"Gr\u00fcnberg","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Ahmed","family":"Abogosh","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"David Gomez","family":"Cabrero","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Stefan T.","family":"Arold","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Robert","family":"Hoehndorf","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Jesper","family":"Tegner","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Narsis A.","family":"Kiani","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]}],"member":"297","published-online":{"date-parts":[[2026,4,27]]},"reference":[{"key":"1167_CR1","doi-asserted-by":"publisher","first-page":"1129","DOI":"10.1016\/j.neuron.2018.08.011","volume":"99","author":"SA Hollingsworth","year":"2018","unstructured":"Hollingsworth SA, Dror RO (2018) Molecular dynamics simulation for all. Neuron 99:1129\u20131143","journal-title":"Neuron"},{"key":"1167_CR2","doi-asserted-by":"publisher","DOI":"10.1016\/j.sbi.2024.102870","volume":"87","author":"T Ginex","year":"2024","unstructured":"Ginex T, V\u00e1zquez J, Estarellas C, Luque FJ (2024) Quantum mechanical-based strategies in drug discovery: finding the pace to new challenges in drug design. Curr Opin Struct Biol 87:102870","journal-title":"Curr Opin Struct Biol"},{"key":"1167_CR3","unstructured":"Zhang X, Wu J, Xie J, Chen T, Zhou K (2024) Benchmark on drug target interaction modeling from a structure perspective. https:\/\/arxiv.org\/abs\/2407.04055v1"},{"key":"1167_CR4","doi-asserted-by":"publisher","DOI":"10.1038\/s41467-025-57828-0","author":"Z Lu","year":"2025","unstructured":"Lu Z et al (2025) Dtiam: a unified framework for predicting drug-target interactions, binding affinities and drug mechanisms. Nat Commun. https:\/\/doi.org\/10.1038\/s41467-025-57828-0","journal-title":"Nat Commun"},{"key":"1167_CR5","doi-asserted-by":"publisher","DOI":"10.1038\/s41467-025-62235-6","author":"Y Zhao","year":"2025","unstructured":"Zhao Y et al (2025) Evidential deep learning-based drug-target interaction prediction. Nat Commun. https:\/\/doi.org\/10.1038\/s41467-025-62235-6","journal-title":"Nat Commun"},{"key":"1167_CR6","doi-asserted-by":"crossref","unstructured":"Liao Q et al (2025) Application of artificial intelligence in drug-target interactions prediction: a review. NPJ Biomed Innov 2","DOI":"10.1038\/s44385-024-00003-9"},{"key":"1167_CR7","doi-asserted-by":"publisher","first-page":"850","DOI":"10.1126\/science.ads0018","volume":"387","author":"T Hayes","year":"2025","unstructured":"Hayes T et al (2025) Simulating 500 million years of evolution with a language model. Science 387:850\u2013858","journal-title":"Science"},{"key":"1167_CR8","unstructured":"Chithrananda S, Grand G, Ramsundar B (2020) ChemBERTa: large-scale self-supervised pretraining for molecular property prediction. arXiv:2010.09885"},{"key":"1167_CR9","doi-asserted-by":"publisher","DOI":"10.1016\/j.crtox.2023.100121","volume":"5","author":"E Ylip\u00e4\u00e4","year":"2023","unstructured":"Ylip\u00e4\u00e4 E et al (2023) Herg-toxicity prediction using traditional machine learning and advanced deep learning techniques. Curr Res Toxicol 5:100121","journal-title":"Curr Res Toxicol"},{"key":"1167_CR10","doi-asserted-by":"publisher","DOI":"10.1101\/684662","author":"T Nguyen","year":"2019","unstructured":"Nguyen T, Le H, Venkatesh S (2019) Graphdta: prediction of drugtarget binding affinity using graph convolutional networks. bioRxiv. https:\/\/doi.org\/10.1101\/684662","journal-title":"bioRxiv"},{"key":"1167_CR11","doi-asserted-by":"publisher","DOI":"10.1016\/j.compbiomed.2023.107621","author":"Z Zhu","year":"2023","unstructured":"Zhu Z et al (2023) Drug\u2013target affinity prediction method based on multi-scale information interaction and graph optimization. Comput Biol Med. https:\/\/doi.org\/10.1016\/j.compbiomed.2023.107621","journal-title":"Comput Biol Med"},{"key":"1167_CR12","doi-asserted-by":"publisher","first-page":"27177","DOI":"10.1007\/s10489-023-04977-8","volume":"53","author":"Y Pan","year":"2023","unstructured":"Pan Y, Zhang Y, Zhang J, Lu M (2023) CSDTI: an interpretable cross-attention network with GNN-based drug molecule aggregation for drug-target interaction prediction. Appl Intell 53:27177\u201327190","journal-title":"Appl Intell"},{"key":"1167_CR13","doi-asserted-by":"publisher","DOI":"10.1016\/j.eswa.2023.121882","author":"J Zhang","year":"2024","unstructured":"Zhang J, Liu Z, Pan Y, Lin H, Zhang Y (2024) IMAEN: an interpretable molecular augmentation model for drug\u2013target interaction prediction. Expert Syst Appl. https:\/\/doi.org\/10.1016\/j.eswa.2023.121882","journal-title":"Expert Syst Appl"},{"key":"1167_CR14","unstructured":"RDKit: Open-Source Cheminformatics Software. https:\/\/www.rdkit.org\/"},{"key":"1167_CR15","doi-asserted-by":"publisher","unstructured":"Shi Y et al (2020) Masked label prediction: unified message passing model for semi-supervised classification. https:\/\/doi.org\/10.48550\/ARXIV.2009.03509","DOI":"10.48550\/ARXIV.2009.03509"},{"key":"1167_CR16","unstructured":"Brody S, Alon U, Yahav E (2022) How attentive are graph attention networks? In: ICLR 2022\u201410th international conference on learning representations (international conference on learning representations. ICLR"},{"key":"1167_CR17","doi-asserted-by":"publisher","DOI":"10.1038\/nbt.1990","author":"MI Davis","year":"2011","unstructured":"Davis MI, Hunt JPHS et al (2011) Comprehensive analysis of kinase inhibitor selectivity. Nat Biotechnol. https:\/\/doi.org\/10.1038\/nbt.1990","journal-title":"Nat Biotechnol"},{"key":"1167_CR18","doi-asserted-by":"publisher","first-page":"735","DOI":"10.1021\/ci400709d","volume":"54","author":"J Tang","year":"2014","unstructured":"Tang J et al (2014) Making sense of large-scale kinase inhibitor bioactivity data sets: a comparative and integrative analysis. J Chem Inf Model 54:735\u2013743","journal-title":"J Chem Inf Model"},{"key":"1167_CR19","doi-asserted-by":"crossref","unstructured":"Tang J, Ravikumar B, Alam Z, Rebane A, V\u00e4h\u00e4-Koskela M, Peddinti G, van Adrichem AJ, Wakkinen J, Jaiswal A, Karjalainen E, Gautam P (2018) Drug target com-mons: a community effort to build a consensus knowl-edge base for drug-target interactions. Cell Chem Biol 25","DOI":"10.1016\/j.chembiol.2017.11.009"},{"key":"1167_CR20","doi-asserted-by":"publisher","DOI":"10.1038\/nchembio.530","author":"J Metz","year":"2011","unstructured":"Metz J et al (2011) Navigating the kinome. Nat Chem Biol. https:\/\/doi.org\/10.1038\/nchembio.530","journal-title":"Nat Chem Biol"},{"key":"1167_CR21","unstructured":"USEPA (2018) US, EPA summary files from invitrodb v2, accessed on 2018-03-22. https:\/\/www.epa.gov\/chemical-research\/toxicity-forecaster-toxcasttm-data"},{"key":"1167_CR22","doi-asserted-by":"publisher","DOI":"10.1093\/nar\/gkm795","author":"M Kuhn","year":"2007","unstructured":"Kuhn M, von Mering C, Campillos M, Jensen LJ, Bork P (2007) Stitch: interaction networks of chemicals and proteins. Nucleic Acids Res. https:\/\/doi.org\/10.1093\/nar\/gkm795","journal-title":"Nucleic Acids Res"},{"key":"1167_CR23","unstructured":"Dasatinib. https:\/\/www.selleckchem.com\/products\/dasatinib-tyrosine-kinase-inhibitor.html."},{"key":"1167_CR24","unstructured":"UM-164. https:\/\/www.selleckchem.com\/products\/um-164.html."},{"key":"1167_CR25","unstructured":"Bafetinib. https:\/\/www.selleckchem.com\/products\/Bafetinib.html."},{"key":"1167_CR26","doi-asserted-by":"crossref","unstructured":"Tsuchiya, Y. et al. Probing strigolactone receptors in Striga hermonthica with fluorescence. Science (1979). 349, 864\u2013868 (2015).","DOI":"10.1126\/science.aab3831"},{"key":"1167_CR27","doi-asserted-by":"crossref","unstructured":"Zhao, Z. et al. Structural Design of Engineered Costimulation Determines Tumor Rejection Kinetics and Persistence of CAR T Cells. Cancer Cell 28, 415\u2013428 (2015).","DOI":"10.1016\/j.ccell.2015.09.004"},{"key":"1167_CR28","unstructured":"Src I1. https:\/\/www.fishersci.com\/shop\/products\/src-i1-2\/364210R."},{"key":"1167_CR29","unstructured":"SKI-1 (ab120839). https:\/\/www.abcam.com\/en-us\/products\/biochemicals\/src-kinase-inhibitor-i-ski-1-ab120839."},{"key":"1167_CR30","unstructured":"AZD7762.https:\/\/www.selleckchem.com\/products\/AZD7762.html"},{"key":"1167_CR31","doi-asserted-by":"crossref","unstructured":"Shahul Hameed, U. et al. Structural basis for specifi c inhibition of the highly sensitive ShHTL7 receptor. EMBORep. 19, (2018).","DOI":"10.15252\/embr.201745619"},{"key":"1167_CR32","unstructured":"Saracatinib. https:\/\/www.selleckchem.com\/products\/AZD0530.html"},{"key":"1167_CR33","unstructured":"WH-4-023. https:\/\/www.abcam.com\/en-us\/products\/biochemicals\/wh-4-023-ab282423."},{"key":"1167_CR34","unstructured":"Veli\u010dkovi\u0107, P. et al. (2017). Graph attention networks. arXiv preprint arXiv:1710.10903 ."},{"key":"1167_CR35","doi-asserted-by":"crossref","unstructured":"Akiba, T., Sano, S., Yanase, T., Ohta, T. & Koyama, M. Optuna (2019). A next-generation hyperparameter optimizationframework. in Proceedings of the 25th ACM SIGKDD international conference on knowledge discovery & data mining2623\u20132631 .","DOI":"10.1145\/3292500.3330701"},{"key":"1167_CR36","doi-asserted-by":"publisher","first-page":"1071","DOI":"10.1002\/jcc.23231","volume":"34","author":"K Roy","year":"2013","unstructured":"Roy K et al (2013) Some case studies on application of , \u00c4\u00farm2, \u00c4\u00f9 metrics for judging quality of quantitative structure, \u00c4\u00ecactivity relationship predictions: emphasis on scaling of response data. J Comput Chem 34:1071\u20131082","journal-title":"J Comput Chem"},{"key":"1167_CR37","doi-asserted-by":"publisher","unstructured":"Barker SC Kassel DB, W. D. H. X. L. M. A. K. W. B. Characterization of PP60C-src tyrosine kinase activities usinga continuous assay: Autoactivation of the enzyme is an intermolecular autophosphorylation process. Biochemistry (1995)https:\/\/doi.org\/10.1021\/bi00045a027","DOI":"10.1021\/bi00045a027"},{"key":"1167_CR38","doi-asserted-by":"publisher","unstructured":"Yung-Chi Cheng & Prusoff , W. H. (1973). Relationship between the inhibition constant (KI) and the concentration ofinhibitor which causes 50 percent inhibition (i50) of an enzymatic reaction. Biochem. Pharmacol. https:\/\/doi.org\/10.1016\/0006-2952(73)90196-2.","DOI":"10.1016\/0006-2952(73)90196-2"}],"container-title":["Journal of Cheminformatics"],"original-title":[],"language":"en","link":[{"URL":"https:\/\/link.springer.com\/article\/10.1186\/s13321-026-01167-9","content-type":"text\/html","content-version":"vor","intended-application":"text-mining"},{"URL":"https:\/\/link.springer.com\/content\/pdf\/10.1186\/s13321-026-01167-9.pdf","content-type":"application\/pdf","content-version":"vor","intended-application":"text-mining"},{"URL":"https:\/\/link.springer.com\/content\/pdf\/10.1186\/s13321-026-01167-9.pdf","content-type":"application\/pdf","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2026,7,23]],"date-time":"2026-07-23T22:46:12Z","timestamp":1784846772000},"score":1,"resource":{"primary":{"URL":"https:\/\/link.springer.com\/10.1186\/s13321-026-01167-9"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2026,4,27]]},"references-count":38,"journal-issue":{"issue":"1","published-online":{"date-parts":[[2026,12]]}},"alternative-id":["1167"],"URL":"https:\/\/doi.org\/10.1186\/s13321-026-01167-9","relation":{},"ISSN":["1758-2946"],"issn-type":[{"value":"1758-2946","type":"electronic"}],"subject":[],"published":{"date-parts":[[2026,4,27]]},"assertion":[{"value":"27 January 2025","order":1,"name":"received","label":"Received","group":{"name":"ArticleHistory","label":"Article History"}},{"value":"2 February 2026","order":2,"name":"accepted","label":"Accepted","group":{"name":"ArticleHistory","label":"Article History"}},{"value":"27 April 2026","order":3,"name":"first_online","label":"First Online","group":{"name":"ArticleHistory","label":"Article History"}},{"value":"The authors declare no competing interests.","order":1,"name":"Ethics","label":"Competing interests","group":{"name":"EthicsHeading","label":"Declarations"}}],"article-number":"101"}}