{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2026,9,14]],"date-time":"2026-09-14T14:18:50Z","timestamp":1789395530550,"version":"build-2803163510"},"reference-count":51,"publisher":"American Chemical Society (ACS)","issue":"17","license":[{"start":{"date-parts":[[2026,8,19]],"date-time":"2026-08-19T00:00:00Z","timestamp":1787097600000},"content-version":"stm-asf","delay-in-days":0,"URL":"https:\/\/doi.org\/10.15223\/policy-029"},{"start":{"date-parts":[[2026,8,19]],"date-time":"2026-08-19T00:00:00Z","timestamp":1787097600000},"content-version":"stm-asf","delay-in-days":0,"URL":"https:\/\/doi.org\/10.15223\/policy-037"},{"start":{"date-parts":[[2026,8,19]],"date-time":"2026-08-19T00:00:00Z","timestamp":1787097600000},"content-version":"stm-asf","delay-in-days":0,"URL":"https:\/\/doi.org\/10.15223\/policy-045"}],"funder":[{"DOI":"10.13039\/na","name":"Guizhou Provincial Graduate Research Fund Project","doi-asserted-by":"publisher","award":["2025YJSKYJJ129"],"award-info":[{"award-number":["2025YJSKYJJ129"]}],"id":[{"id":"10.13039\/na","id-type":"DOI","asserted-by":"publisher"}]},{"DOI":"10.13039\/na","name":"National Natural Science Foundation of China","doi-asserted-by":"publisher","award":["32260225"],"award-info":[{"award-number":["32260225"]}],"id":[{"id":"10.13039\/na","id-type":"DOI","asserted-by":"publisher"}]}],"content-domain":{"domain":[],"crossmark-restriction":false},"short-container-title":[],"published-print":{"date-parts":[[2026,9,14]]},"abstract":"<jats:title>Abstract<\/jats:title>\n                  <jats:p>Accurate prediction of drug\u2013protein interactions (DPIs) is crucial for accelerating the drug discovery process. However, the scarcity of experimentally validated interactions can limit the learning of transferable interaction patterns, particularly for previously unseen drugs and proteins. To address this fundamental challenge, we propose the MMU-DPI framework. A key component of this framework is a Label Mix strategy tailored to multimodal DPI prediction, which performs interpolation only in the label space while keeping the input modalities unchanged. This strategy provides stochastic soft-target regularization and improves generalization performance under reduced-data and independent external Cold-both evaluation settings. To effectively process and utilize multimodal data, MMU-DPI adopts a multimodal dual-branch architecture. The first branch uses a Message Passing Neural Network (MPNN) to extract structured representations from drug molecular graphs. It also uses a Convolutional Neural Network (CNN) to capture key biological and functional features from amino acid sequences. The second branch constructs a heterogeneous interaction graph and uses a Graph Attention Network (GAT) to learn deep contextual relationships between drugs and proteins. A learnable global fusion weight combines complementary branch logits to generate the final prediction for each drug\u2013protein pair. Experimental results on multiple benchmark data sets demonstrate that MMU-DPI outperforms several state-of-the-art DPI prediction methods. Case studies further support the ability of MMU-DPI to identify potential DPIs. These results indicate that MMU-DPI can serve as a useful computational tool for drug discovery.<\/jats:p>","DOI":"10.1021\/acs.jcim.6c01188","type":"journal-article","created":{"date-parts":[[2026,8,19]],"date-time":"2026-08-19T16:01:16Z","timestamp":1787155276000},"page":"10548-10563","source":"Crossref","is-referenced-by-count":0,"title":["MMU-DPI: Enhancing\nGeneralization in Drug\u2013Protein\nInteraction Prediction through Multimodal Learning and a Label Mix\nStrategy"],"prefix":"10.1021","volume":"66","author":[{"ORCID":"https:\/\/orcid.org\/0009-0007-8693-6086","authenticated-orcid":true,"given":"Jiahao","family":"Wei","sequence":"first","affiliation":[{"name":"Guizhou Normal University , , ,","place":["Guiyang, China, 550025"]}],"role":[{"vocabulary":"crossref","role":"author"}]},{"ORCID":"https:\/\/orcid.org\/0000-0001-7313-3494","authenticated-orcid":true,"given":"Tie","family":"Shen","sequence":"additional","affiliation":[{"name":"Guizhou Normal University , , ,","place":["Guiyang, China, 550001"]}],"role":[{"vocabulary":"crossref","role":"author"}]}],"member":"316","published-online":{"date-parts":[[2026,8,19]]},"reference":[{"issue":"7403","key":"2026091409065819100_cit1","doi-asserted-by":"publisher","first-page":"361","DOI":"10.1038\/nature11159","article-title":"Large-scale prediction\nand testing of drug activity on side-effect targets","volume":"486","author":"Lounkine","year":"2012","journal-title":"Nature"},{"issue":"7270","key":"2026091409065819100_cit2","doi-asserted-by":"publisher","first-page":"175","DOI":"10.1038\/nature08506","article-title":"Predicting new molecular\ntargets for known drugs","volume":"462","author":"Keiser","year":"2009","journal-title":"Nature"},{"issue":"7","key":"2026091409065819100_cit3","doi-asserted-by":"publisher","first-page":"361","DOI":"10.1089\/omi.2022.0068","article-title":"Multi-omics and artificial\nintelligence-guided drug repositioning:\nProspects, challenges, and lessons learned from COVID-19","volume":"26","author":"Cong","year":"2022","journal-title":"OMICS"},{"issue":"1","key":"2026091409065819100_cit4","doi-asserted-by":"publisher","first-page":"451","DOI":"10.1093\/bib\/bbz152","article-title":"DTI-CDF: A cascade deep\nforest model towards the prediction of drug\u2013target interactions\nbased on hybrid features","volume":"22","author":"Chu","year":"2021","journal-title":"Brief. Bioinform."},{"issue":"21","key":"2026091409065819100_cit5","doi-asserted-by":"publisher","first-page":"1421","DOI":"10.1016\/S1359-6446(05)03632-9","article-title":"Keynote review: in vitro\nsafety pharmacology profiling: an essential tool for successful drug\ndevelopment","volume":"10","author":"Whitebread","year":"2005","journal-title":"Drug Discovery Today"},{"issue":"1","key":"2026091409065819100_cit6","doi-asserted-by":"publisher","first-page":"71","DOI":"10.1038\/nbt1273","article-title":"Structure-based maximal\naffinity model predicts small-molecule druggability","volume":"25","author":"Cheng","year":"2007","journal-title":"Nat. Biotechnol."},{"issue":"2","key":"2026091409065819100_cit7","doi-asserted-by":"publisher","first-page":"197","DOI":"10.1038\/nbt1284","article-title":"Relating protein pharmacology\nby ligand chemistry","volume":"25","author":"Keiser","year":"2007","journal-title":"Nat. Biotechnol."},{"issue":"5","key":"2026091409065819100_cit8","doi-asserted-by":"publisher","first-page":"905","DOI":"10.1038\/nprot.2016.051","article-title":"Computational protein\u2013ligand\ndocking and virtual drug screening with the AutoDock suite","volume":"11","author":"Forli","year":"2016","journal-title":"Nat. Protoc."},{"issue":"1","key":"2026091409065819100_cit9","doi-asserted-by":"publisher","first-page":"247","DOI":"10.1093\/bib\/bbz157","article-title":"Machine learning approaches\nand databases for prediction of drug\u2013target interaction: a\nsurvey paper","volume":"22","author":"Bagherian","year":"2021","journal-title":"Brief. Bioinform."},{"issue":"5","key":"2026091409065819100_cit10","doi-asserted-by":"publisher","first-page":"1350","DOI":"10.1016\/j.drudis.2022.02.023","article-title":"Compound\u2013protein interaction prediction\nby deep learning: databases, descriptors and models","volume":"27","author":"Du","year":"2022","journal-title":"Drug Discovery Today"},{"issue":"17","key":"2026091409065819100_cit11","doi-asserted-by":"publisher","first-page":"i821","DOI":"10.1093\/bioinformatics\/bty593","article-title":"DeepDTA: deep\ndrug-target binding\naffinity prediction","volume":"34","author":"O\u0308ztu\u0308rk","year":"2018","journal-title":"Bioinformatics"},{"issue":"6","key":"2026091409065819100_cit12","doi-asserted-by":"publisher","first-page":"830","DOI":"10.1093\/bioinformatics\/btaa880","article-title":"MolTrans: molecular\ninteraction transformer for drug\u2013target interaction prediction","volume":"37","author":"Huang","year":"2021","journal-title":"Bioinformatics"},{"issue":"8","key":"2026091409065819100_cit13","doi-asserted-by":"publisher","first-page":"1140","DOI":"10.1093\/bioinformatics\/btaa921","article-title":"GraphDTA:\nPredicting Drug\u2013Target Binding Affinity with Graph Neural Networks","volume":"37","author":"Nguyen","year":"2021","journal-title":"Bioinformatics"},{"issue":"2","key":"2026091409065819100_cit14","doi-asserted-by":"publisher","first-page":"bbad082","DOI":"10.1093\/bib\/bbad082","article-title":"MCANet: shared-weight-based\nMultiheadCrossAttention network for drug\u2013target interaction\nprediction","volume":"24","author":"Bian","year":"2023","journal-title":"Brief. Bioinform."},{"issue":"4","key":"2026091409065819100_cit15","doi-asserted-by":"publisher","first-page":"bbac269","DOI":"10.1093\/bib\/bbac269","article-title":"Mitigating cold-start\nproblems in\ndrug\u2013target affinity prediction with interaction knowledge\ntransferring","volume":"23","author":"Nguyen","year":"2022","journal-title":"Brief. Bioinform."},{"issue":"1","key":"2026091409065819100_cit16","doi-asserted-by":"publisher","first-page":"285","DOI":"10.1109\/TCBB.2022.3144008","article-title":"CPInformer for efficient\nand robust compound\u2013protein interaction prediction","volume":"20","author":"Hua","year":"2023","journal-title":"IEEE\/ACM Trans. Comput. Biol. Bioinform."},{"key":"2026091409065819100_cit17","doi-asserted-by":"publisher","first-page":"11106","DOI":"10.1609\/aaai.v35i12.17325","article-title":"Informer: Beyond efficient\ntransformer for long sequence time-series forecasting","volume":"35","author":"Zhou","year":"2021","journal-title":"Proc. AAAI Conf. Artif. Intell. Online"},{"issue":"2","key":"2026091409065819100_cit18","doi-asserted-by":"publisher","first-page":"btad056","DOI":"10.1093\/bioinformatics\/btad056","article-title":"MFR-DTA: a multi-functional\nand robust model for predicting drug\u2013target binding affinity\nand region","volume":"39","author":"Hua","year":"2023","journal-title":"Bioinformatics"},{"key":"2026091409065819100_cit19","doi-asserted-by":"crossref","unstructured":"Wu, L.; Huang, Y.; Tan, C.; Gao, Z.; Hu, B.; Lin, H.; Liu, Z.; Li, S. Z.\n          PSC-CPI: Multi-scale\nprotein sequence\u2013structure contrasting for efficient and generalizable\ncompound\u2013protein interaction prediction;  Proceedings of the AAAI Conference on Artificial Intelligence; Association for the Advancement of Artificial Intelligence\n(AAAI), Vancouver, Canada, 2024; Vol. 38, pp 310\u2013319.","DOI":"10.1609\/aaai.v38i1.27784"},{"issue":"12","key":"2026091409065819100_cit20","doi-asserted-by":"publisher","first-page":"btae693","DOI":"10.1093\/bioinformatics\/btae693","article-title":"Accurate and transferable drug\u2013target\ninteraction prediction with DrugLAMP","volume":"40","author":"Luo","year":"2024","journal-title":"Bioinformatics"},{"issue":"1","key":"2026091409065819100_cit21","doi-asserted-by":"publisher","first-page":"711","DOI":"10.1109\/JBHI.2024.3476120","article-title":"DTI-MvSCA: An anti-over-smoothing multi-view\nframework with negative sample selection for predicting drug\u2013target\ninteractions","volume":"29","author":"Peng","year":"2025","journal-title":"IEEE J. Biomed. Health Inform."},{"issue":"3","key":"2026091409065819100_cit22","doi-asserted-by":"publisher","first-page":"1318","DOI":"10.1021\/acs.jcim.4c02088","article-title":"DO-GMA: An end-to-end drug\u2013target interaction\nidentification framework with a depthwise overparameterized convolutional\nnetwork and the gated multihead attention mechanism","volume":"65","author":"Peng","year":"2025","journal-title":"J. Chem. Inf. Model."},{"issue":"16","key":"2026091409065819100_cit23","doi-asserted-by":"publisher","first-page":"8871","DOI":"10.1021\/acs.jcim.5c01142","article-title":"HMBVIP: A\nNovel Hierarchical Multi-Bio-View Intelligent Prediction Networks\nfor Drug\u2013Target Interaction Prediction","volume":"65","author":"Yang","year":"2025","journal-title":"J. Chem. Inf. Model."},{"key":"2026091409065819100_cit24","doi-asserted-by":"crossref","unstructured":"Hua, Y.; Xu, T.; Song, X.; Feng, Z.; Wang, R.; Zhang, W.; Wu, X.\n          R-DTI:\nDrug\u2013target interaction prediction based on second-order relevance\nexploration;  Proceedings of the AAAI Conference\non Artificial Intelligence; Association for\nthe Advancement of Artificial Intelligence (AAAI), Philadelphia, PA, USA, 2025; Vol. 39, pp 17368\u201317376.","DOI":"10.1609\/aaai.v39i16.33909"},{"issue":"6","key":"2026091409065819100_cit25","doi-asserted-by":"publisher","first-page":"1239","DOI":"10.1111\/j.1476-5381.2010.01127.x","article-title":"Principles of early\ndrug discovery","volume":"162","author":"Hughes","year":"2011","journal-title":"Br. J. Pharmacol."},{"issue":"2","key":"2026091409065819100_cit26","doi-asserted-by":"publisher","first-page":"325","DOI":"10.1093\/bib\/bbu010","article-title":"Toward more realistic\ndrug\u2013target interaction predictions","volume":"16","author":"Pahikkala","year":"2015","journal-title":"Brief. Bioinform."},{"issue":"D1","key":"2026091409065819100_cit27","doi-asserted-by":"publisher","first-page":"D1045","DOI":"10.1093\/nar\/gkv1072","article-title":"BindingDB in 2015: a public database for\nmedicinal chemistry, computational chemistry and systems pharmacology","volume":"44","author":"Gilson","year":"2016","journal-title":"Nucleic Acids Res."},{"issue":"11","key":"2026091409065819100_cit28","doi-asserted-by":"publisher","first-page":"1046","DOI":"10.1038\/nbt.1990","article-title":"Comprehensive analysis\nof kinase inhibitor selectivity","volume":"29","author":"Davis","year":"2011","journal-title":"Nat. Biotechnol."},{"issue":"3","key":"2026091409065819100_cit29","doi-asserted-by":"publisher","first-page":"735","DOI":"10.1021\/ci400709d","article-title":"Making sense of large-scale\nkinase inhibitor bioactivity data sets: a comparative and integrative\nanalysis","volume":"54","author":"Tang","year":"2014","journal-title":"J. Chem. Inf. Model."},{"issue":"1","key":"2026091409065819100_cit30","doi-asserted-by":"publisher","first-page":"573","DOI":"10.1038\/s41467-017-00680-8","article-title":"A network\nintegration approach for drug\u2013target interaction prediction\nand computational drug repositioning from heterogeneous information","volume":"8","author":"Luo","year":"2017","journal-title":"Nat. Commun."},{"issue":"15","key":"2026091409065819100_cit31","doi-asserted-by":"publisher","first-page":"6582","DOI":"10.1021\/jm300687e","article-title":"Directory of useful\ndecoys, enhanced (DUD-E): Better ligands and decoys for better benchmarking","volume":"55","author":"Mysinger","year":"2012","journal-title":"J. Med. Chem."},{"issue":"4","key":"2026091409065819100_cit32","doi-asserted-by":"publisher","first-page":"942","DOI":"10.1021\/acs.jcim.6b00740","article-title":"Protein\u2013ligand\nscoring with convolutional neural networks","volume":"57","author":"Ragoza","year":"2017","journal-title":"J. Chem. Inf. Model."},{"issue":"D1","key":"2026091409065819100_cit33","doi-asserted-by":"publisher","first-page":"D1265","DOI":"10.1093\/nar\/gkad976","article-title":"DrugBank\n6.0: the DrugBank\nKnowledgebase for 2024","volume":"52","author":"Knox","year":"2024","journal-title":"Nucleic Acids Res."},{"key":"2026091409065819100_cit34","doi-asserted-by":"publisher","first-page":"807","DOI":"10.1007\/s12539-025-00774-8","article-title":"A scalable and robust ensemble deep\nlearning method for predicting drug\u2013target interactions","volume":"18","author":"Cheng","year":"2026","journal-title":"Interdiscip. Sci. Comput. Life Sci."},{"issue":"D1","key":"2026091409065819100_cit35","doi-asserted-by":"publisher","first-page":"D1372","DOI":"10.1093\/nar\/gkae791","article-title":"DrugMAP 2.0: molecular atlas and pharma-information\nof all drugs","volume":"53","author":"Li","year":"2025","journal-title":"Nucleic Acids Res."},{"key":"2026091409065819100_cit36","unstructured":"Gilmer, J.; Schoenholz, S. S.; et al.; Neural Message Passing for Quantum\nChemistry;  Proceedings of Machine Learning\nResearch; PMLR, Sydney, NSW, Australia, 2017; Vol. 70, pp 1263\u20131272."},{"key":"2026091409065819100_cit37","doi-asserted-by":"publisher","first-page":"1281880","DOI":"10.3389\/fmed.2023.1281880","article-title":"Accurately\nidentifying hemagglutinin\nusing sequence information and machine learning methods","volume":"10","author":"Zou","year":"2023","journal-title":"Front. Med."},{"issue":"14","key":"2026091409065819100_cit38","doi-asserted-by":"publisher","first-page":"2465","DOI":"10.3390\/diagnostics13142465","article-title":"A first computational\nframe for recognizing heparin-binding protein","volume":"13","author":"Zhu","year":"2023","journal-title":"Diagnostics"},{"key":"2026091409065819100_cit39","unstructured":"Velic\u030ckovic\u0301, P.; Cucurull, G.; Casanova, A.; Romero, A.; Lio\u0300, P.; Bengio, Y.\n          Graph\nAttention Networks. In  Proceedings of Machine\nLearning Research; Vancouver: BC, Canada, 2018."},{"key":"2026091409065819100_cit40","unstructured":"Abu-El-Haija, S.; Perozzi, B.; Kapoor, A.; et al.; Mixhop: Higher-Order Graph Convolutional Architectures via Sparsified\nNeighborhood Mixing;  Proceedings of Machine\nLearning Research; PMLR, Long Beach, California, USA, 2019; Vol. 97, pp 21\u201329."},{"key":"2026091409065819100_cit41","unstructured":"Zhang, H.; Cisse\u0301, M.; Dauphin, Y. N.; et al.; mixup: Beyond empirical risk minimization. In  Proceedings of Machine Learning Research; Vancouver, BC, Canada, 2018."},{"key":"2026091409065819100_cit42","doi-asserted-by":"publisher","first-page":"12775","DOI":"10.1021\/acs.jcim.5c02054","article-title":"Robust Chemical\nReaction Condition\nRecommendations via Label Mix Strategy","volume":"65","author":"Yan","year":"2025","journal-title":"J. Chem.\nInf. Model."},{"issue":"17","key":"2026091409065819100_cit43","doi-asserted-by":"publisher","first-page":"2664","DOI":"10.1093\/bioinformatics\/btw228","article-title":"Drug repositioning based\non comprehensive similarity measures and Bi-Random walk algorithm","volume":"32","author":"Luo","year":"2016","journal-title":"Bioinformatics"},{"issue":"2","key":"2026091409065819100_cit44","first-page":"627","article-title":"Receiver operating\ncharacteristic (ROC) curve analysis for medical diagnostic test evaluation","volume":"4","author":"Hajian-Tilaki","year":"2013","journal-title":"Caspian J. Int. Med."},{"key":"2026091409065819100_cit45","doi-asserted-by":"crossref","unstructured":"Davis, J.; Goadrich, M.\n          The Relationship\nBetween Precision\u2013Recall and ROC Curves;  Proceedings of the 23rd international conference on Machine learning\n-ICML \u201906; ACM Press, Pittsburgh, Pennsylvania, USA, 2006; Vol. 148, pp 233\u2013240.","DOI":"10.1145\/1143844.1143874"},{"issue":"5","key":"2026091409065819100_cit46","doi-asserted-by":"publisher","first-page":"1946","DOI":"10.1109\/TCBB.2020.2964221","article-title":"iLncRNAdis-FB:\nidentify lncRNA\u2013disease\nassociations by fusing biological feature blocks through deep neural\nnetwork","volume":"18","author":"Wei","year":"2021","journal-title":"IEEE\/ACM Trans. Comput. Biol. Bioinform."},{"issue":"3","key":"2026091409065819100_cit47","doi-asserted-by":"publisher","first-page":"1602","DOI":"10.1109\/JBHI.2024.3375025","article-title":"BINDTI: A bi-directional intention network\nfor drug\u2013target interaction identification based on attention\nmechanisms","volume":"29","author":"Peng","year":"2025","journal-title":"IEEE J. Biomed. Health Inform"},{"issue":"16","key":"2026091409065819100_cit48","doi-asserted-by":"publisher","first-page":"6684","DOI":"10.1021\/acs.jcim.4c00957","article-title":"MGNDTI: A drug\u2013target interaction prediction\nframework based on multimodal representation learning and the gating\nmechanism","volume":"64","author":"Peng","year":"2024","journal-title":"J. Chem. Inf. Model."},{"issue":"D1","key":"2026091409065819100_cit49","doi-asserted-by":"publisher","first-page":"D1074","DOI":"10.1093\/nar\/gkx1037","article-title":"DrugBank 5.0: a major\nupdate to the DrugBank database for 2018","volume":"46","author":"Wishart","year":"2018","journal-title":"Nucleic\nAcids Res."},{"issue":"D1","key":"2026091409065819100_cit50","doi-asserted-by":"publisher","first-page":"D963","DOI":"10.1093\/nar\/gky963","article-title":"DrugCentral 2018: an\nupdate","volume":"47","author":"Ursu","year":"2019","journal-title":"Nucleic Acids Res."},{"issue":"D1","key":"2026091409065819100_cit51","doi-asserted-by":"publisher","first-page":"D506","DOI":"10.1093\/nar\/gky1049","article-title":"UniProt:\na worldwide hub of protein knowledge","volume":"47","author":"The UniProt Consortium.","year":"2019","journal-title":"Nucleic\nAcids Res."}],"container-title":["Journal of Chemical Information\nand Modeling"],"original-title":[],"language":"en","link":[{"URL":"https:\/\/pubs.acs.org\/doi\/pdf\/10.1021\/acs.jcim.6c01188","content-type":"application\/pdf","content-version":"vor","intended-application":"unspecified"},{"URL":"https:\/\/pubs.acs.org\/jcisd8\/article-pdf\/66\/17\/10548\/66881906\/acs.jcim.6c01188.pdf","content-type":"application\/pdf","content-version":"vor","intended-application":"syndication"},{"URL":"https:\/\/pubs.acs.org\/jcisd8\/article-pdf\/66\/17\/10548\/66881906\/acs.jcim.6c01188.pdf","content-type":"unspecified","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2026,9,14]],"date-time":"2026-09-14T13:22:47Z","timestamp":1789392167000},"score":1,"resource":{"primary":{"URL":"https:\/\/pubs.acs.org\/jcisd8\/article\/66\/17\/10548\/5276883\/MMU-DPI-Enhancing-Generalization-in-Drug-Protein"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2026,8,19]]},"references-count":51,"journal-issue":{"issue":"17","published-online":{"date-parts":[[2026,8,19]]},"published-print":{"date-parts":[[2026,9,14]]}},"URL":"https:\/\/doi.org\/10.1021\/acs.jcim.6c01188","relation":{},"ISSN":["1549-9596","1549-960X"],"issn-type":[{"value":"1549-9596","type":"print"},{"value":"1549-960X","type":"electronic"}],"subject":[],"published":{"date-parts":[[2026,8,19]]}}}