{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2026,8,17]],"date-time":"2026-08-17T23:14:42Z","timestamp":1787008482452,"version":"build-2736575974"},"reference-count":39,"publisher":"Oxford University Press (OUP)","issue":"2","license":[{"start":{"date-parts":[[2018,7,6]],"date-time":"2018-07-06T00:00:00Z","timestamp":1530835200000},"content-version":"vor","delay-in-days":0,"URL":"https:\/\/academic.oup.com\/journals\/pages\/open_access\/funder_policies\/chorus\/standard_publication_model"}],"funder":[{"DOI":"10.13039\/501100001863","name":"NEDO","doi-asserted-by":"publisher","id":[{"id":"10.13039\/501100001863","id-type":"DOI","asserted-by":"publisher"}]},{"name":"JSPS KAKENHI","award":["JP17H07392"],"award-info":[{"award-number":["JP17H07392"]}]},{"name":"Platform Project for Supporting Drug Discovery and Life Science Research"},{"name":"Basis for Supporting Innovative Drug Discovery and Life Science Research"},{"name":"BINDS"},{"DOI":"10.13039\/100009619","name":"AMED","doi-asserted-by":"publisher","award":["JP18am0101110"],"award-info":[{"award-number":["JP18am0101110"]}],"id":[{"id":"10.13039\/100009619","id-type":"DOI","asserted-by":"publisher"}]},{"name":"JST CREST","award":["JPMJCR1689"],"award-info":[{"award-number":["JPMJCR1689"]}]},{"name":"JSPS KAKENHI","award":["15H01717"],"award-info":[{"award-number":["15H01717"]}]}],"content-domain":{"domain":[],"crossmark-restriction":false},"short-container-title":[],"published-print":{"date-parts":[[2019,1,15]]},"abstract":"<jats:title>Abstract<\/jats:title>\n                  <jats:sec>\n                    <jats:title>Motivation<\/jats:title>\n                    <jats:p>In bioinformatics, machine learning-based methods that predict the compound\u2013protein interactions (CPIs) play an important role in the virtual screening for drug discovery. Recently, end-to-end representation learning for discrete symbolic data (e.g. words in natural language processing) using deep neural networks has demonstrated excellent performance on various difficult problems. For the CPI problem, data are provided as discrete symbolic data, i.e. compounds are represented as graphs where the vertices are atoms, the edges are chemical bonds, and proteins are sequences in which the characters are amino acids. In this study, we investigate the use of end-to-end representation learning for compounds and proteins, integrate the representations, and develop a new CPI prediction approach by combining a graph neural network (GNN) for compounds and a convolutional neural network (CNN) for proteins.<\/jats:p>\n                  <\/jats:sec>\n                  <jats:sec>\n                    <jats:title>Results<\/jats:title>\n                    <jats:p>Our experiments using three CPI datasets demonstrated that the proposed end-to-end approach achieves competitive or higher performance as compared to various existing CPI prediction methods. In addition, the proposed approach significantly outperformed existing methods on an unbalanced dataset. This suggests that data-driven representations of compounds and proteins obtained by end-to-end GNNs and CNNs are more robust than traditional chemical and biological features obtained from databases. Although analyzing deep learning models is difficult due to their black-box nature, we address this issue using a neural attention mechanism, which allows us to consider which subsequences in a protein are more important for a drug compound when predicting its interaction. The neural attention mechanism also provides effective visualization, which makes it easier to analyze a model even when modeling is performed using real-valued representations instead of discrete features.<\/jats:p>\n                  <\/jats:sec>\n                  <jats:sec>\n                    <jats:title>Availability and implementation<\/jats:title>\n                    <jats:p>https:\/\/github.com\/masashitsubaki<\/jats:p>\n                  <\/jats:sec>\n                  <jats:sec>\n                    <jats:title>Supplementary information<\/jats:title>\n                    <jats:p>Supplementary data are available at Bioinformatics online.<\/jats:p>\n                  <\/jats:sec>","DOI":"10.1093\/bioinformatics\/bty535","type":"journal-article","created":{"date-parts":[[2018,7,4]],"date-time":"2018-07-04T14:53:11Z","timestamp":1530715991000},"page":"309-318","source":"Crossref","is-referenced-by-count":641,"title":["Compound\u2013protein interaction prediction with end-to-end learning of neural networks for graphs and sequences"],"prefix":"10.1093","volume":"35","author":[{"given":"Masashi","family":"Tsubaki","sequence":"first","affiliation":[{"name":"National Institute of Advanced Industrial Science and Technology, Artificial Intelligence Research Center, Tokyo, Japan"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Kentaro","family":"Tomii","sequence":"additional","affiliation":[{"name":"National Institute of Advanced Industrial Science and Technology, Artificial Intelligence Research Center, Tokyo, Japan"},{"name":"AIST- Tokyo Tech Real World Big-Data Computation Open Innovation Laboratory, Tokyo, Japan"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Jun","family":"Sese","sequence":"additional","affiliation":[{"name":"National Institute of Advanced Industrial Science and Technology, Artificial Intelligence Research Center, Tokyo, Japan"},{"name":"AIST- Tokyo Tech Real World Big-Data Computation Open Innovation Laboratory, Tokyo, Japan"}],"role":[{"vocabulary":"crossref","role":"author"}]}],"member":"286","published-online":{"date-parts":[[2018,7,6]]},"reference":[{"key":"2023013107231513600_bty535-B1","doi-asserted-by":"crossref","first-page":"55827","DOI":"10.1074\/jbc.M409792200","article-title":"A novel mode of gleevec binding is revealed by the structure of spleen tyrosine kinase","volume":"279","author":"Atwell","year":"2004","journal-title":"J. Biol. Chem"},{"key":"2023013107231513600_bty535-B2","author":"Bahdanau","year":"2014"},{"key":"2023013107231513600_bty535-B3","doi-asserted-by":"crossref","first-page":"2397","DOI":"10.1093\/bioinformatics\/btp433","article-title":"Supervised prediction of drug\u2013target interactions using bipartite local models","volume":"25","author":"Bleakley","year":"2009","journal-title":"Bioinformatics"},{"key":"2023013107231513600_bty535-B4","doi-asserted-by":"crossref","first-page":"262","DOI":"10.1038\/nrg1317","article-title":"Chemogenomics: an emerging strategy for rapid target and drug discovery","volume":"5","author":"Bredel","year":"2004","journal-title":"Nat. Rev. Genet"},{"key":"2023013107231513600_bty535-B5","doi-asserted-by":"crossref","first-page":"2373","DOI":"10.1039\/c2mb25110h","article-title":"Prediction of chemical\u2013protein interactions: multitarget-qsar versus computational chemogenomic methods","volume":"8","author":"Cheng","year":"2012","journal-title":"Mol. BioSyst"},{"key":"2023013107231513600_bty535-B6","author":"Costa","year":"2010"},{"key":"2023013107231513600_bty535-B7","doi-asserted-by":"crossref","first-page":"734","DOI":"10.1093\/bib\/bbt056","article-title":"Similarity-based machine learning methods for predicting drug\u2013target interactions: a brief review","volume":"15","author":"Ding","year":"2014","journal-title":"Brief. Bioinf"},{"key":"2023013107231513600_bty535-B8","doi-asserted-by":"crossref","first-page":"285","DOI":"10.1093\/bioinformatics\/bti801","article-title":"Application of latent semantic analysis to protein remote homology detection","volume":"22","author":"Dong","year":"2006","journal-title":"Bioinformatics"},{"key":"2023013107231513600_bty535-B9","author":"Gilmer","year":"2017"},{"key":"2023013107231513600_bty535-B10","doi-asserted-by":"crossref","first-page":"2304","DOI":"10.1093\/bioinformatics\/bts360","article-title":"Predicting drug\u2013target interactions from chemical and genomic kernels using Bayesian matrix factorization","volume":"28","author":"G\u00f6nen","year":"2012","journal-title":"Bioinformatics"},{"key":"2023013107231513600_bty535-B11","doi-asserted-by":"crossref","first-page":"D919","DOI":"10.1093\/nar\/gkm862","article-title":"Supertarget and matador: resources for exploring drug-target relationships","volume":"36","author":"G\u00fcnther","year":"2008","journal-title":"Nucleic Acids Res"},{"key":"2023013107231513600_bty535-B12","doi-asserted-by":"crossref","first-page":"1600045.","DOI":"10.1002\/minf.201600045","article-title":"Cgbvs-dnn: prediction of compound\u2013protein interactions based on deep learning","volume":"36","author":"Hamanaka","year":"2017","journal-title":"Mol. Inf"},{"key":"2023013107231513600_bty535-B13","doi-asserted-by":"crossref","first-page":"82","DOI":"10.1109\/MSP.2012.2205597","article-title":"Deep neural networks for acoustic modeling in speech recognition: the shared views of four research groups","volume":"29","author":"Hinton","year":"2012","journal-title":"IEEE Signal Process. Mag"},{"key":"2023013107231513600_bty535-B14","doi-asserted-by":"crossref","first-page":"2149","DOI":"10.1093\/bioinformatics\/btn409","article-title":"Protein-ligand interaction prediction: an improved chemogenomics approach","volume":"24","author":"Jacob","year":"2008","journal-title":"Bioinformatics"},{"key":"2023013107231513600_bty535-B15","doi-asserted-by":"crossref","first-page":"595","DOI":"10.1007\/s10822-016-9938-8","article-title":"Molecular graph convolutions: moving beyond fingerprints","volume":"30","author":"Kearnes","year":"2016","journal-title":"J. Comput. Aided Mol. Des"},{"key":"2023013107231513600_bty535-B16","doi-asserted-by":"crossref","first-page":"175","DOI":"10.1038\/nature08506","article-title":"Predicting new molecular targets for known drugs","volume":"462","author":"Keiser","year":"2009","journal-title":"Nature"},{"key":"2023013107231513600_bty535-B17","first-page":"5882","article-title":"Convolutional neural networks for sentence classification","volume":"1408","author":"Kim","year":"2014","journal-title":"arXiv Preprint arXiv"},{"key":"2023013107231513600_bty535-B18","author":"Kingma","year":"2014"},{"key":"2023013107231513600_bty535-B19","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":"2023013107231513600_bty535-B20","first-page":"1097","author":"Krizhevsky","year":"2012"},{"key":"2023013107231513600_bty535-B21","doi-asserted-by":"crossref","first-page":"436","DOI":"10.1038\/nature14539","article-title":"Deep learning","volume":"521","author":"LeCun","year":"2015","journal-title":"Nature"},{"key":"2023013107231513600_bty535-B22","doi-asserted-by":"crossref","first-page":"i221","DOI":"10.1093\/bioinformatics\/btv256","article-title":"Improving compound\u2013protein interaction prediction by building up highly credible negative samples","volume":"31","author":"Liu","year":"2015","journal-title":"Bioinformatics"},{"key":"2023013107231513600_bty535-B23","doi-asserted-by":"crossref","first-page":"361","DOI":"10.1038\/nature11159","article-title":"Large-scale prediction and testing of drug activity on side-effect targets","volume":"486","author":"Lounkine","year":"2012","journal-title":"Nature"},{"key":"2023013107231513600_bty535-B24","doi-asserted-by":"crossref","first-page":"495","DOI":"10.1016\/j.drudis.2013.01.008","article-title":"Shifting from the single to the multitarget paradigm in drug discovery","volume":"18","author":"Medina-Franco","year":"2013","journal-title":"Drug Discovery Today"},{"key":"2023013107231513600_bty535-B25","doi-asserted-by":"crossref","first-page":"6582","DOI":"10.1021\/jm300687e","article-title":"Directory of useful decoys, enhanced (dud-e): better ligands and decoys for better benchmarking","volume":"55","author":"Mysinger","year":"2012","journal-title":"J. Med. Chem"},{"key":"2023013107231513600_bty535-B26","doi-asserted-by":"crossref","first-page":"942","DOI":"10.1021\/acs.jcim.6b00740","article-title":"Protein\u2013ligand scoring with convolutional neural networks","volume":"57","author":"Ragoza","year":"2017","journal-title":"J. Chem. Inf. Model"},{"key":"2023013107231513600_bty535-B27","doi-asserted-by":"crossref","first-page":"61","DOI":"10.1109\/TNN.2008.2005605","article-title":"The graph neural network model","volume":"20","author":"Scarselli","year":"2009","journal-title":"IEEE Trans. Neural Netw"},{"key":"2023013107231513600_bty535-B28","doi-asserted-by":"crossref","first-page":"13890.","DOI":"10.1038\/ncomms13890","article-title":"Quantum-chemical insights from deep tensor neural networks","volume":"8","author":"Sch\u00fctt","year":"2017","journal-title":"Nat. Commun"},{"key":"2023013107231513600_bty535-B29","doi-asserted-by":"crossref","first-page":"757","DOI":"10.2174\/156802607780487669","article-title":"Structural elements of ligand recognition site in secretory phospholipase a2 and structure-based design of specific inhibitors","volume":"7","author":"Singh","year":"2007","journal-title":"Curr. Top. Med. Chem"},{"key":"2023013107231513600_bty535-B30","first-page":"3104","author":"Sutskever","year":"2014"},{"key":"2023013107231513600_bty535-B31","doi-asserted-by":"crossref","first-page":"S3","DOI":"10.1186\/1752-0509-7-S6-S3","article-title":"Scalable prediction of compound\u2013protein interactions using minwise hashing","volume":"7","author":"Tabei","year":"2013","journal-title":"BMC Syst. Biol"},{"key":"2023013107231513600_bty535-B32","doi-asserted-by":"crossref","first-page":"64","DOI":"10.1016\/j.ymeth.2016.06.024","article-title":"Boosting compound\u2013protein interaction prediction by deep learning","volume":"110","author":"Tian","year":"2016","journal-title":"Methods"},{"key":"2023013107231513600_bty535-B33","author":"Tokui","year":"2015"},{"key":"2023013107231513600_bty535-B34","doi-asserted-by":"crossref","first-page":"NA","DOI":"10.1002\/jcc.21334","article-title":"Autodock vina: improving the speed and accuracy of docking with a new scoring function, efficient optimization, and multithreading","volume":"31","author":"Trott","year":"2009","journal-title":"J. Comput. Chem"},{"key":"2023013107231513600_bty535-B35","doi-asserted-by":"crossref","first-page":"3036","DOI":"10.1093\/bioinformatics\/btr500","article-title":"Gaussian interaction profile kernels for predicting drug\u2013target interaction","volume":"27","author":"van Laarhoven","year":"2011","journal-title":"Bioinformatics"},{"key":"2023013107231513600_bty535-B36","first-page":"02855","article-title":"Atomnet: a deep convolutional neural network for bioactivity prediction in structure-based drug discovery","volume":"1510","author":"Wallach","year":"2015","journal-title":"arXiv Preprint arXiv"},{"key":"2023013107231513600_bty535-B37","first-page":"086033","article-title":"Deep learning with feature embedding for compound\u2013protein interaction prediction","author":"Wan","year":"2016","journal-title":"bioRxiv"},{"key":"2023013107231513600_bty535-B38","doi-asserted-by":"crossref","first-page":"D901","DOI":"10.1093\/nar\/gkm958","article-title":"Drugbank: a knowledgebase for drugs, drug actions and drug targets","volume":"36","author":"Wishart","year":"2008","journal-title":"Nucleic Acids Res"},{"key":"2023013107231513600_bty535-B39","doi-asserted-by":"crossref","first-page":"i232","DOI":"10.1093\/bioinformatics\/btn162","article-title":"Prediction of drug\u2013target interaction networks from the integration of chemical and genomic spaces","volume":"24","author":"Yamanishi","year":"2008","journal-title":"Bioinformatics"}],"container-title":["Bioinformatics"],"original-title":[],"language":"en","link":[{"URL":"https:\/\/academic.oup.com\/bioinformatics\/article-pdf\/35\/2\/309\/48963189\/bioinformatics_35_2_309.pdf","content-type":"application\/pdf","content-version":"vor","intended-application":"syndication"},{"URL":"https:\/\/academic.oup.com\/bioinformatics\/article-pdf\/35\/2\/309\/48963189\/bioinformatics_35_2_309.pdf","content-type":"unspecified","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2023,1,31]],"date-time":"2023-01-31T05:08:33Z","timestamp":1675141713000},"score":1,"resource":{"primary":{"URL":"https:\/\/academic.oup.com\/bioinformatics\/article\/35\/2\/309\/5050020"}},"subtitle":[],"editor":[{"given":"Jonathan","family":"Wren","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"editor"}]}],"short-title":[],"issued":{"date-parts":[[2018,7,6]]},"references-count":39,"journal-issue":{"issue":"2","published-print":{"date-parts":[[2019,1,15]]}},"URL":"https:\/\/doi.org\/10.1093\/bioinformatics\/bty535","relation":{},"ISSN":["1367-4803","1367-4811"],"issn-type":[{"value":"1367-4803","type":"print"},{"value":"1367-4811","type":"electronic"}],"subject":[],"published-other":{"date-parts":[[2019,1,15]]},"published":{"date-parts":[[2018,7,6]]}}}