{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2025,12,8]],"date-time":"2025-12-08T17:14:26Z","timestamp":1765214066933,"version":"3.46.0"},"reference-count":42,"publisher":"Oxford University Press (OUP)","issue":"12","license":[{"start":{"date-parts":[[2025,11,9]],"date-time":"2025-11-09T00:00:00Z","timestamp":1762646400000},"content-version":"vor","delay-in-days":0,"URL":"https:\/\/creativecommons.org\/licenses\/by\/4.0\/"}],"funder":[{"name":"National Natural Science Foundation of China Original Exploration","award":["62450112"],"award-info":[{"award-number":["62450112"]}]},{"DOI":"10.13039\/501100001809","name":"National Natural Science Foundation of China","doi-asserted-by":"publisher","award":["32470687"],"award-info":[{"award-number":["32470687"]}],"id":[{"id":"10.13039\/501100001809","id-type":"DOI","asserted-by":"publisher"}]},{"DOI":"10.13039\/501100012166","name":"National Key R& D Program of China","doi-asserted-by":"crossref","award":["2024YFF1206602"],"award-info":[{"award-number":["2024YFF1206602"]}],"id":[{"id":"10.13039\/501100012166","id-type":"DOI","asserted-by":"crossref"}]},{"DOI":"10.13039\/501100012226","name":"Fundamental Research Funds for the Central Universities","doi-asserted-by":"publisher","award":["2572025JT05"],"award-info":[{"award-number":["2572025JT05"]}],"id":[{"id":"10.13039\/501100012226","id-type":"DOI","asserted-by":"publisher"}]},{"name":"Key Project of the Natural Science Foundation of Heilongjiang Province","award":["ZD2024F001"],"award-info":[{"award-number":["ZD2024F001"]}]}],"content-domain":{"domain":[],"crossmark-restriction":false},"short-container-title":[],"published-print":{"date-parts":[[2025,12,1]]},"abstract":"<jats:title>Abstract<\/jats:title>\n                  <jats:sec>\n                    <jats:title>Motivation<\/jats:title>\n                    <jats:p>Accurately predicting protein-ligand interactions is fundamental to elucidating molecular recognition and has far-reaching implications in drug discovery, gene regulation, and signal transduction. Conventional methods predominantly rely on internal structural or sequence-based protein representations. While these approaches have improved predictive performance, their dependence on limited labeled data restricts the capacity to learn expressive features from structural inputs. Moreover, they often neglect the intricate geometric and chemical context encoded on protein surfaces, limiting interpretability, and hindering mechanistic insights into binding interactions.<\/jats:p>\n                  <\/jats:sec>\n                  <jats:sec>\n                    <jats:title>Result<\/jats:title>\n                    <jats:p>Here, we present PLiSAGE, a multimodal framework that integrates 3D structural and surface geometric embeddings to enable accurate prediction of protein\u2013ligand interactions. Central to our approach is the joint pretraining of structural and surface encoders through unsupervised contrastive learning and point cloud reconstruction. Protein surfaces are represented as segmented point cloud patches, allowing the model to capture fine-grained geometric and chemical cues. A Transformer-based encoder further captures both local and global spatial dependencies across patches. The incorporation of spatial topological information during pretraining facilitates the learning of stable, discriminative, and multi-scale protein representations, enhancing the expressive capacity of both modalities. An adaptive fusion module dynamically integrates structural and surface embeddings to yield complete and robust protein representations. PLiSAGE demonstrates superior performance over competitive baselines in binding affinity prediction and interaction classification tasks. Extensive ablation studies underscore the critical contributions of surface features and the pretraining strategy to the model\u2019s generalization capabilities.<\/jats:p>\n                  <\/jats:sec>\n                  <jats:sec>\n                    <jats:title>Availability and implementation<\/jats:title>\n                    <jats:p>The source code of PLiSAGE is available at: https:\/\/github.com\/catly\/PLiSAGE.<\/jats:p>\n                  <\/jats:sec>","DOI":"10.1093\/bioinformatics\/btaf608","type":"journal-article","created":{"date-parts":[[2025,11,9]],"date-time":"2025-11-09T20:35:02Z","timestamp":1762720502000},"source":"Crossref","is-referenced-by-count":0,"title":["PLiSAGE: enhancing protein-ligand interaction prediction with multimodal surface and geometry encoding"],"prefix":"10.1093","volume":"41","author":[{"given":"Tianci","family":"Wang","sequence":"first","affiliation":[{"name":"College of Computer and Control Engineering, Northeast Forestry University , Harbin, 150040,","place":["China"]}]},{"ORCID":"https:\/\/orcid.org\/0009-0004-8128-0047","authenticated-orcid":false,"given":"Guanyu","family":"Qiao","sequence":"additional","affiliation":[{"name":"Faculty of Computing, Harbin Institute of Technology , Harbin, 150001,","place":["China"]}]},{"ORCID":"https:\/\/orcid.org\/0000-0001-7381-2374","authenticated-orcid":false,"given":"Guohua","family":"Wang","sequence":"additional","affiliation":[{"name":"Faculty of Computing, Harbin Institute of Technology , Harbin, 150001,","place":["China"]}]},{"ORCID":"https:\/\/orcid.org\/0000-0002-0403-7287","authenticated-orcid":false,"given":"Yang","family":"Li","sequence":"additional","affiliation":[{"name":"College of Computer and Control Engineering, Northeast Forestry University , Harbin, 150040,","place":["China"]}]}],"member":"286","published-online":{"date-parts":[[2025,11,9]]},"reference":[{"first-page":"4","year":"2022","author":"Aykent","key":"2025120812120689000_btaf608-B1"},{"key":"2025120812120689000_btaf608-B2","doi-asserted-by":"crossref","first-page":"235","DOI":"10.1145\/357306.357310","article-title":"A generalization of algebraic surface drawing","volume":"1","author":"Blinn","year":"1982","journal-title":"ACM Trans Graph"},{"key":"2025120812120689000_btaf608-B3","first-page":"26","article-title":"Efficient curvature estimation for oriented point clouds","volume":"1050","author":"Cao","year":"2019","journal-title":"Stat"},{"key":"2025120812120689000_btaf608-B4","doi-asserted-by":"crossref","first-page":"1046","DOI":"10.1038\/nbt.1990","article-title":"Comprehensive analysis of kinase inhibitor selectivity","volume":"29","author":"Davis","year":"2011","journal-title":"Nat Biotechnol"},{"key":"2025120812120689000_btaf608-B5","first-page":"82","article-title":"Pymol: an open-source molecular graphics tool","volume":"40","author":"DeLano","year":"2002","journal-title":"CCP4 Newsl Protein Crystallogr"},{"first-page":"605","year":"2017","author":"Fan","key":"2025120812120689000_btaf608-B6"},{"key":"2025120812120689000_btaf608-B7","doi-asserted-by":"crossref","first-page":"830","DOI":"10.1093\/bioinformatics\/btaa880","article-title":"Moltrans: molecular interaction transformer for drug\u2013target interaction prediction","volume":"37","author":"Huang","year":"2021","journal-title":"Bioinformatics"},{"key":"2025120812120689000_btaf608-B8","doi-asserted-by":"crossref","first-page":"bbad464","DOI":"10.1093\/bib\/bbad464","article-title":"Thitogene: a deep learning method for predicting spatial transcriptomics from histological images","volume":"25","author":"Jia","year":"2023","journal-title":"Brief Bioinform"},{"key":"2025120812120689000_btaf608-B9","doi-asserted-by":"crossref","first-page":"D1270","DOI":"10.1093\/nar\/gkae837","article-title":"Xqtlatlas: a comprehensive resource for human cellular-resolution multi-omics genetic regulatory landscape","volume":"53","author":"Jia","year":"2025","journal-title":"Nucleic Acids Res"},{"key":"2025120812120689000_btaf608-B10","doi-asserted-by":"crossref","first-page":"18209","DOI":"10.1021\/acs.jmedchem.1c01830","article-title":"Interactiongraphnet: a novel and efficient deep graph representation learning framework for accurate protein\u2013ligand interaction predictions","volume":"64","author":"Jiang","year":"2021","journal-title":"J Med Chem"},{"author":"Jing","key":"2025120812120689000_btaf608-B11"},{"key":"2025120812120689000_btaf608-B12","doi-asserted-by":"crossref","first-page":"1710","DOI":"10.3390\/pharmaceutics14081710","article-title":"Fine-tuning of bert model to accurately predict drug\u2013target interactions","volume":"14","author":"Kang","year":"2022","journal-title":"Pharmaceutics"},{"key":"2025120812120689000_btaf608-B13","doi-asserted-by":"crossref","first-page":"1893","DOI":"10.1021\/ci300604z","article-title":"Lessons learned in empirical scoring with smina from the CSAR 2011 benchmarking exercise","volume":"53","author":"Koes","year":"2013","journal-title":"J Chem Inf Model"},{"key":"2025120812120689000_btaf608-B14","doi-asserted-by":"crossref","first-page":"e1007129","DOI":"10.1371\/journal.pcbi.1007129","article-title":"DeepConv-DTI: prediction of drug-target interactions via deep learning with convolution on protein sequences","volume":"15","author":"Lee","year":"2019","journal-title":"PLoS Comput Biol"},{"key":"2025120812120689000_btaf608-B15","doi-asserted-by":"crossref","first-page":"14","DOI":"10.1186\/s13321-024-00808-1","article-title":"DLM-DTI: a dual language model for the prediction of drug-target interaction with hint-based learning","volume":"16","author":"Lee","year":"2024","journal-title":"J Cheminform"},{"key":"2025120812120689000_btaf608-B16","doi-asserted-by":"crossref","first-page":"e60\u2013e60","DOI":"10.1093\/nar\/gkad288","article-title":"GeoBind: segmentation of nucleic acid binding interface on protein surface with geometric deep learning","volume":"51","author":"Li","year":"2023","journal-title":"Nucleic Acids Res"},{"key":"2025120812120689000_btaf608-B17","doi-asserted-by":"crossref","first-page":"2847","DOI":"10.1093\/bioinformatics\/btac164","article-title":"Supervised graph co-contrastive learning for drug\u2013target interaction prediction","volume":"38","author":"Li","year":"2022","journal-title":"Bioinformatics"},{"author":"Liang","key":"2025120812120689000_btaf608-B18"},{"key":"2025120812120689000_btaf608-B19","doi-asserted-by":"crossref","first-page":"405","DOI":"10.1093\/bioinformatics\/btu626","article-title":"Pdb-wide collection of binding data: current status of the pdbbind database","volume":"31","author":"Liu","year":"2015","journal-title":"Bioinformatics"},{"key":"2025120812120689000_btaf608-B20","first-page":"7236","article-title":"TANKBind: trigonometry-aware neural networks for drug-protein binding structure prediction","volume":"35","author":"Lu","year":"2022","journal-title":"Adv Neural Inf Process Syst"},{"key":"2025120812120689000_btaf608-B21","doi-asserted-by":"crossref","first-page":"1427","DOI":"10.1038\/s42256-023-00756-9","article-title":"Learning characteristics of graph neural networks predicting protein\u2013ligand affinities","volume":"5","author":"Mastropietro","year":"2023","journal-title":"Nat Mach Intell"},{"key":"2025120812120689000_btaf608-B22","doi-asserted-by":"crossref","first-page":"43","DOI":"10.1186\/s13321-021-00522-2","article-title":"GNINA 1.0: molecular docking with deep learning","volume":"13","author":"McNutt","year":"2021","journal-title":"J Cheminform"},{"key":"2025120812120689000_btaf608-B23","doi-asserted-by":"crossref","first-page":"1140","DOI":"10.1093\/bioinformatics\/btaa921","article-title":"GraphDTA: predicting drug\u2013target binding affinity with graph neural networks","volume":"37","author":"Nguyen","year":"2021","journal-title":"Bioinformatics"},{"key":"2025120812120689000_btaf608-B24","doi-asserted-by":"crossref","first-page":"bpae043","DOI":"10.1093\/biomethods\/bpae043","article-title":"Multimodal pretraining for unsupervised protein representation learning","volume":"9","author":"Nguyen","year":"2024","journal-title":"Biol Methods Protoc"},{"key":"2025120812120689000_btaf608-B25","doi-asserted-by":"crossref","first-page":"i821","DOI":"10.1093\/bioinformatics\/bty593","article-title":"DeepDTA: deep drug\u2013target binding affinity prediction","volume":"34","author":"\u00d6zt\u00fcrk","year":"2018","journal-title":"Bioinformatics"},{"key":"2025120812120689000_btaf608-B26","doi-asserted-by":"crossref","first-page":"1414","DOI":"10.1080\/07391102.2023.2291829","article-title":"DeepCompoundNet: enhancing compound\u2013protein interaction prediction with multimodal convolutional neural networks","volume":"43","author":"Palhamkhani","year":"2025","journal-title":"J Biomol Struct Dyn"},{"first-page":"604","year":"2022","author":"Pang","key":"2025120812120689000_btaf608-B27"},{"first-page":"652","year":"2017","author":"Qi","key":"2025120812120689000_btaf608-B28"},{"key":"2025120812120689000_btaf608-B29","doi-asserted-by":"crossref","first-page":"btae570","DOI":"10.1093\/bioinformatics\/btae570","article-title":"Causal enhanced drug\u2013target interaction prediction based on graph generation and multi-source information fusion","volume":"40","author":"Qiao","year":"2024","journal-title":"Bioinformatics"},{"key":"2025120812120689000_btaf608-B30","doi-asserted-by":"crossref","first-page":"8","DOI":"10.1136\/mp.53.1.8","article-title":"X ray crystallography","volume":"53","author":"Smyth","year":"2000","journal-title":"Mol Pathol"},{"key":"2025120812120689000_btaf608-B31","doi-asserted-by":"crossref","first-page":"3666","DOI":"10.1093\/bioinformatics\/bty374","article-title":"Development and evaluation of a deep learning model for protein\u2013ligand binding affinity prediction","volume":"34","author":"Stepniewska-Dziubinska","year":"2018","journal-title":"Bioinformatics"},{"key":"2025120812120689000_btaf608-B32","doi-asserted-by":"crossref","first-page":"895","DOI":"10.1021\/acs.jcim.8b00545","article-title":"Comparative assessment of scoring functions: the CASF-2016 update","volume":"59","author":"Su","year":"2019","journal-title":"J Chem Inf Model"},{"first-page":"15272","year":"2021","author":"Sverrisson","key":"2025120812120689000_btaf608-B33"},{"key":"2025120812120689000_btaf608-B34","doi-asserted-by":"crossref","first-page":"D480","DOI":"10.1093\/nar\/gkaa1100","article-title":"Uniprot: the universal protein knowledgebase in 2021","volume":"49","year":"2021","journal-title":"Nucleic Acids Res"},{"key":"2025120812120689000_btaf608-B35","doi-asserted-by":"crossref","first-page":"D368","DOI":"10.1093\/nar\/gkad1011","article-title":"Alphafold protein structure database in 2024: providing structure coverage for over 214 million protein sequences","volume":"52","author":"Varadi","year":"2024","journal-title":"Nucleic Acids Res"},{"key":"2025120812120689000_btaf608-B36","doi-asserted-by":"crossref","first-page":"D668","DOI":"10.1093\/nar\/gkj067","article-title":"DrugBank: a comprehensive resource for in silico drug discovery and exploration","volume":"34","author":"Wishart","year":"2006","journal-title":"Nucleic Acids Res"},{"key":"2025120812120689000_btaf608-B37","doi-asserted-by":"crossref","first-page":"btae413","DOI":"10.1093\/bioinformatics\/btae413","article-title":"Surface-based multimodal protein\u2013ligand binding affinity prediction","volume":"40","author":"Xu","year":"2024","journal-title":"Bioinformatics"},{"key":"2025120812120689000_btaf608-B38","doi-asserted-by":"crossref","first-page":"e324","DOI":"10.1371\/journal.pbio.0030324","article-title":"Design of wide-spectrum inhibitors targeting coronavirus main proteases","volume":"3","author":"Yang","year":"2005","journal-title":"PLoS Biol"},{"first-page":"19313","year":"2022","author":"Yu","key":"2025120812120689000_btaf608-B39"},{"key":"2025120812120689000_btaf608-B40","doi-asserted-by":"crossref","first-page":"4691","DOI":"10.3390\/molecules28124691","article-title":"DeepBindGCN: integrating molecular vector representation with graph convolutional neural networks for protein\u2013ligand interaction prediction","volume":"28","author":"Zhang","year":"2023","journal-title":"Molecules"},{"key":"2025120812120689000_btaf608-B41","doi-asserted-by":"crossref","first-page":"4295","DOI":"10.1109\/JBHI.2024.3384238","article-title":"A hierarchical graph neural network framework for predicting protein-protein interaction modulators with functional group information and hypergraph structure","volume":"28","author":"Zhang","year":"2024","journal-title":"IEEE J Biomed Health Inform"},{"key":"2025120812120689000_btaf608-B42","doi-asserted-by":"crossref","first-page":"103227","DOI":"10.1016\/j.inffus.2025.103227","article-title":"Improving protein-protein interaction modulator predictions via knowledge-fused language models","volume":"123","author":"Zhang","year":"2025","journal-title":"Inform Fusion"}],"container-title":["Bioinformatics"],"original-title":[],"language":"en","link":[{"URL":"https:\/\/academic.oup.com\/bioinformatics\/advance-article-pdf\/doi\/10.1093\/bioinformatics\/btaf608\/65248500\/btaf608.pdf","content-type":"application\/pdf","content-version":"am","intended-application":"syndication"},{"URL":"https:\/\/academic.oup.com\/bioinformatics\/article-pdf\/41\/12\/btaf608\/65248500\/btaf608.pdf","content-type":"application\/pdf","content-version":"vor","intended-application":"syndication"},{"URL":"https:\/\/academic.oup.com\/bioinformatics\/article-pdf\/41\/12\/btaf608\/65248500\/btaf608.pdf","content-type":"unspecified","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2025,12,8]],"date-time":"2025-12-08T17:12:18Z","timestamp":1765213938000},"score":1,"resource":{"primary":{"URL":"https:\/\/academic.oup.com\/bioinformatics\/article\/doi\/10.1093\/bioinformatics\/btaf608\/8317435"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2025,11,9]]},"references-count":42,"journal-issue":{"issue":"12","published-print":{"date-parts":[[2025,12,1]]}},"URL":"https:\/\/doi.org\/10.1093\/bioinformatics\/btaf608","relation":{},"ISSN":["1367-4803","1367-4811"],"issn-type":[{"type":"print","value":"1367-4803"},{"type":"electronic","value":"1367-4811"}],"subject":[],"published-other":{"date-parts":[[2025,12]]},"published":{"date-parts":[[2025,11,9]]},"article-number":"btaf608"}}