{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2026,4,17]],"date-time":"2026-04-17T17:44:48Z","timestamp":1776447888882,"version":"3.51.2"},"reference-count":61,"publisher":"Oxford University Press (OUP)","issue":"13","license":[{"start":{"date-parts":[[2018,6,27]],"date-time":"2018-06-27T00:00:00Z","timestamp":1530057600000},"content-version":"vor","delay-in-days":0,"URL":"http:\/\/creativecommons.org\/licenses\/by-nc\/4.0\/"}],"funder":[{"DOI":"10.13039\/100000001","name":"NSF","doi-asserted-by":"publisher","award":["IIS-1149837"],"award-info":[{"award-number":["IIS-1149837"]}],"id":[{"id":"10.13039\/100000001","id-type":"DOI","asserted-by":"publisher"}]},{"DOI":"10.13039\/100000002","name":"NIH","doi-asserted-by":"publisher","award":["BD2K U54EB020405"],"award-info":[{"award-number":["BD2K U54EB020405"]}],"id":[{"id":"10.13039\/100000002","id-type":"DOI","asserted-by":"publisher"}]},{"name":"DARPA SIMPLEX"},{"name":"Stanford Data Science Initiative"},{"name":"Chan Zuckerberg Biohub"}],"content-domain":{"domain":[],"crossmark-restriction":false},"short-container-title":[],"published-print":{"date-parts":[[2018,7,1]]},"abstract":"<jats:title>Abstract<\/jats:title>\n                  <jats:sec>\n                    <jats:title>Motivation<\/jats:title>\n                    <jats:p>The use of drug combinations, termed polypharmacy, is common to treat patients with complex diseases or co-existing conditions. However, a major consequence of polypharmacy is a much higher risk of adverse side effects for the patient. Polypharmacy side effects emerge because of drug\u2013drug interactions, in which activity of one drug may change, favorably or unfavorably, if taken with another drug. The knowledge of drug interactions is often limited because these complex relationships are rare, and are usually not observed in relatively small clinical testing. Discovering polypharmacy side effects thus remains an important challenge with significant implications for patient mortality and morbidity.<\/jats:p>\n                  <\/jats:sec>\n                  <jats:sec>\n                    <jats:title>Results<\/jats:title>\n                    <jats:p>Here, we present Decagon, an approach for modeling polypharmacy side effects. The approach constructs a multimodal graph of protein\u2013protein interactions, drug\u2013protein target interactions and the polypharmacy side effects, which are represented as drug\u2013drug interactions, where each side effect is an edge of a different type. Decagon is developed specifically to handle such multimodal graphs with a large number of edge types. Our approach develops a new graph convolutional neural network for multirelational link prediction in multimodal networks. Unlike approaches limited to predicting simple drug\u2013drug interaction values, Decagon can predict the exact side effect, if any, through which a given drug combination manifests clinically. Decagon accurately predicts polypharmacy side effects, outperforming baselines by up to 69%. We find that it automatically learns representations of side effects indicative of co-occurrence of polypharmacy in patients. Furthermore, Decagon models particularly well polypharmacy side effects that have a strong molecular basis, while on predominantly non-molecular side effects, it achieves good performance because of effective sharing of model parameters across edge types. Decagon opens up opportunities to use large pharmacogenomic and patient population data to flag and prioritize polypharmacy side effects for follow-up analysis via formal pharmacological studies.<\/jats:p>\n                  <\/jats:sec>\n                  <jats:sec>\n                    <jats:title>Availability and implementation<\/jats:title>\n                    <jats:p>Source code and preprocessed datasets are at: http:\/\/snap.stanford.edu\/decagon.<\/jats:p>\n                  <\/jats:sec>","DOI":"10.1093\/bioinformatics\/bty294","type":"journal-article","created":{"date-parts":[[2018,4,12]],"date-time":"2018-04-12T15:32:51Z","timestamp":1523547171000},"page":"i457-i466","source":"Crossref","is-referenced-by-count":1079,"title":["Modeling polypharmacy side effects with graph convolutional networks"],"prefix":"10.1093","volume":"34","author":[{"given":"Marinka","family":"Zitnik","sequence":"first","affiliation":[{"name":"Department of Computer Science, Stanford University, Stanford, CA, USA"}]},{"given":"Monica","family":"Agrawal","sequence":"additional","affiliation":[{"name":"Department of Computer Science, Stanford University, Stanford, CA, USA"}]},{"given":"Jure","family":"Leskovec","sequence":"additional","affiliation":[{"name":"Department of Computer Science, Stanford University, Stanford, CA, USA"},{"name":"Chan Zuckerberg Biohub, San Francisco, CA, USA"}]}],"member":"286","published-online":{"date-parts":[[2018,6,27]]},"reference":[{"key":"2023051605023635700_bty294-B1","doi-asserted-by":"crossref","first-page":"1","DOI":"10.1155\/2017\/3801819","article-title":"Severe rhabdomyolysis due to presumed drug interactions between atorvastatin with amlodipine and ticagrelor","volume":"2017","author":"Banakh","year":"2017","journal-title":"Case Rep. Crit. Care"},{"key":"2023051605023635700_bty294-B2","doi-asserted-by":"crossref","first-page":"1213","DOI":"10.1038\/nbt.3052","article-title":"A community computational challenge to predict the activity of pairs of compounds","volume":"32","author":"Bansal","year":"2014","journal-title":"Nat. Biotechnol"},{"key":"2023051605023635700_bty294-B3","doi-asserted-by":"crossref","first-page":"1282","DOI":"10.1124\/dmd.117.077883","article-title":"Elucidation of the impact of p-glycoprotein and breast cancer resistance protein on the brain distribution of catechol-o-methyltransferase inhibitors","volume":"45","author":"Bicker","year":"2017","journal-title":"Drug Metab. Dispos"},{"key":"2023051605023635700_bty294-B4","doi-asserted-by":"crossref","first-page":"263","DOI":"10.1126\/science.1158140","article-title":"Drug target identification using side-effect similarity","volume":"321","author":"Campillos","year":"2008","journal-title":"Science"},{"key":"2023051605023635700_bty294-B5","doi-asserted-by":"crossref","first-page":"D470","DOI":"10.1093\/nar\/gku1204","article-title":"The BioGRID interaction database: 2015 update","volume":"43","author":"Chatr-Aryamontri","year":"2015","journal-title":"Nucleic Acids Res"},{"key":"2023051605023635700_bty294-B6","doi-asserted-by":"crossref","first-page":"614","DOI":"10.1039\/C5MB00599J","article-title":"Synergy evaluation by a pathway\u2013pathway interaction network: a new way to predict drug combination","volume":"12","author":"Chen","year":"2016","journal-title":"Mol. BioSyst"},{"key":"2023051605023635700_bty294-B7","doi-asserted-by":"crossref","first-page":"e1004975","DOI":"10.1371\/journal.pcbi.1004975","article-title":"NLLSS: predicting synergistic drug combinations based on semi-supervised learning","volume":"12","author":"Chen","year":"2016","journal-title":"PLoS Comput. Biol"},{"key":"2023051605023635700_bty294-B8","doi-asserted-by":"crossref","first-page":"e278","DOI":"10.1136\/amiajnl-2013-002512","article-title":"Machine learning-based prediction of drug\u2013drug interactions by integrating drug phenotypic, therapeutic, chemical, and genomic properties","volume":"21","author":"Cheng","year":"2014","journal-title":"J. Am. Med. Inform. Assoc"},{"key":"2023051605023635700_bty294-B9","author":"Defferrard","year":"2016"},{"key":"2023051605023635700_bty294-B10","first-page":"192","article-title":"Drug-related morbidity and mortality: updating the cost-of-illness model","volume":"41","author":"Ernst","year":"2001","journal-title":"J. Am. Pharm. Assoc"},{"key":"2023051605023635700_bty294-B11","doi-asserted-by":"crossref","first-page":"54","DOI":"10.1016\/j.jbi.2017.04.021","article-title":"Computational prediction of drug-drug interactions based on drugs functional similarities","volume":"70","author":"Ferdousi","year":"2017","journal-title":"J. Biomed. Inform"},{"key":"2023051605023635700_bty294-B12","first-page":"1263","author":"Gilmer","year":"2017"},{"key":"2023051605023635700_bty294-B13","first-page":"249","author":"Glorot","year":"2010"},{"key":"2023051605023635700_bty294-B14","doi-asserted-by":"crossref","first-page":"592.","DOI":"10.1038\/msb.2012.26","article-title":"INDI: a computational framework for inferring drug interactions and their associated recommendations","volume":"8","author":"Gottlieb","year":"2012","journal-title":"Mol. Syst. Biol"},{"key":"2023051605023635700_bty294-B15","first-page":"1025","author":"Hamilton","year":"2017"},{"key":"2023051605023635700_bty294-B16","first-page":"52","author":"Hamilton","year":"2017"},{"key":"2023051605023635700_bty294-B17","first-page":"463","author":"Han","year":"2017"},{"key":"2023051605023635700_bty294-B18","doi-asserted-by":"crossref","first-page":"186","DOI":"10.1002\/wsbm.1337","article-title":"In silico methods for drug repurposing and pharmacology","volume":"8","author":"Hodos","year":"2016","journal-title":"Wiley Interdisc. Rev. Syst. Biol. Med"},{"key":"2023051605023635700_bty294-B19","doi-asserted-by":"crossref","first-page":"i228","DOI":"10.1093\/bioinformatics\/btu278","article-title":"Drugcomboranker: drug combination discovery based on target network analysis","volume":"30","author":"Huang","year":"2014","journal-title":"Bioinformatics"},{"key":"2023051605023635700_bty294-B20","doi-asserted-by":"crossref","DOI":"10.1038\/srep07160","article-title":"Systematic prediction of drug combinations based on clinical side-effects","volume":"4","author":"Huang","year":"2014","journal-title":"Sci. Rep"},{"key":"2023051605023635700_bty294-B21","doi-asserted-by":"crossref","first-page":"111","DOI":"10.1038\/nrd2683","article-title":"Mechanisms of drug combinations: interaction and network perspectives","volume":"8","author":"Jia","year":"2009","journal-title":"Nat. Rev. Drug Discov"},{"key":"2023051605023635700_bty294-B22","doi-asserted-by":"crossref","first-page":"1818","DOI":"10.1001\/jama.2015.13766","article-title":"Trends in prescription drug use among adults in the United States from 1999-2012","volume":"314","author":"Kantor","year":"2015","journal-title":"JAMA"},{"key":"2023051605023635700_bty294-B23","author":"Kingma","year":"2014"},{"key":"2023051605023635700_bty294-B24","author":"Kipf","year":"2016"},{"key":"2023051605023635700_bty294-B25","doi-asserted-by":"crossref","first-page":"D1075","DOI":"10.1093\/nar\/gkv1075","article-title":"The SIDER database of drugs and side effects","volume":"44","author":"Kuhn","year":"2016","journal-title":"Nucleic Acids Res"},{"key":"2023051605023635700_bty294-B26","doi-asserted-by":"crossref","first-page":"9880","DOI":"10.1073\/pnas.0802208105","article-title":"The implications of human metabolic network topology for disease comorbidity","volume":"105","author":"Lee","year":"2008","journal-title":"Proc. Natl. Acad. Sci. USA"},{"key":"2023051605023635700_bty294-B27","doi-asserted-by":"crossref","first-page":"36.","DOI":"10.1186\/s13321-015-0090-6","article-title":"Synergy maps: exploring compound combinations using network-based visualization","volume":"7","author":"Lewis","year":"2015","journal-title":"J. Cheminform"},{"key":"2023051605023635700_bty294-B28","doi-asserted-by":"crossref","first-page":"2","DOI":"10.1093\/bib\/bbv020","article-title":"A survey of current trends in computational drug repositioning","volume":"17","author":"Li","year":"2016","journal-title":"Brief. Bioinform"},{"key":"2023051605023635700_bty294-B29","first-page":"35","author":"Li","year":"2017"},{"key":"2023051605023635700_bty294-B30","doi-asserted-by":"crossref","first-page":"410","DOI":"10.1038\/nrd1720","article-title":"Elucidating mechanisms of drug-induced toxicity","volume":"4","author":"Liebler","year":"2005","journal-title":"Nat. Rev. Drug Discov"},{"key":"2023051605023635700_bty294-B31","first-page":"285","article-title":"The problem of synergism and antagonism of combined drugs","volume":"3","author":"Loewe","year":"1953","journal-title":"Arzneimittel-Forschung"},{"key":"2023051605023635700_bty294-B32","first-page":"2579","article-title":"Visualizing data using t-SNE","volume":"9","author":"Maaten","year":"2008","journal-title":"J. Mach. Learn. Res"},{"key":"2023051605023635700_bty294-B33","doi-asserted-by":"crossref","first-page":"1257601.","DOI":"10.1126\/science.1257601","article-title":"Uncovering disease-disease relationships through the incomplete interactome","volume":"347","author":"Menche","year":"2015","journal-title":"Science"},{"key":"2023051605023635700_bty294-B34","first-page":"3111","author":"Mikolov","year":"2013"},{"key":"2023051605023635700_bty294-B35","author":"Nickel","year":"2011"},{"key":"2023051605023635700_bty294-B36","doi-asserted-by":"crossref","first-page":"748","DOI":"10.1016\/j.ccell.2017.11.003","article-title":"Synthetic lethality of combined Bcl-2 inhibition and p53 activation in AML: mechanisms and superior antileukemic efficacy","volume":"32","author":"Pan","year":"2017","journal-title":"Cancer Cell"},{"key":"2023051605023635700_bty294-B37","doi-asserted-by":"crossref","first-page":"16.","DOI":"10.1145\/2915921","article-title":"Tensors for data mining and data fusion: models, applications, and scalable algorithms","volume":"8","author":"Papalexakis","year":"2017","journal-title":"ACM Trans. Intell. Syst. Technol"},{"key":"2023051605023635700_bty294-B38","doi-asserted-by":"crossref","first-page":"2204","DOI":"10.1056\/NEJMoa1208799","article-title":"Cardiorenal end points in a trial of aliskiren for type 2 diabetes","volume":"367","author":"Parving","year":"2012","journal-title":"N. Engl. J. Med"},{"key":"2023051605023635700_bty294-B39","author":"Percha","year":"2012"},{"key":"2023051605023635700_bty294-B40","author":"Perozzi","year":"2014"},{"key":"2023051605023635700_bty294-B41","doi-asserted-by":"crossref","first-page":"1212","DOI":"10.1016\/j.cell.2014.10.050","article-title":"A proteome-scale map of the human interactome network","volume":"159","author":"Rolland","year":"2014","journal-title":"Cell"},{"key":"2023051605023635700_bty294-B42","doi-asserted-by":"crossref","first-page":"70","DOI":"10.1016\/j.saa.2015.11.021","article-title":"Looking for the interactions between omeprazole and amoxicillin in a disordered phase. an experimental and theoretical study","volume":"156","author":"Russo","year":"2016","journal-title":"Spectrochim. Acta A"},{"key":"2023051605023635700_bty294-B43","doi-asserted-by":"crossref","first-page":"7.","DOI":"10.1186\/s13321-015-0055-9","article-title":"Systems biology approaches for advancing the discovery of effective drug combinations","volume":"7","author":"Ryall","year":"2015","journal-title":"J. Cheminform"},{"key":"2023051605023635700_bty294-B44","author":"Schlichtkrull","year":"2017"},{"key":"2023051605023635700_bty294-B45","doi-asserted-by":"crossref","first-page":"409.","DOI":"10.1186\/s12859-017-1818-2","article-title":"Predicting combinative drug pairs towards realistic screening via integrating heterogeneous features","volume":"18","author":"Shi","year":"2017","journal-title":"BMC Bioinformatics"},{"key":"2023051605023635700_bty294-B46","first-page":"1929","article-title":"Dropout: a simple way to prevent neural networks from overfitting","volume":"15","author":"Srivastava","year":"2014","journal-title":"J. Mach. Learn. Res"},{"key":"2023051605023635700_bty294-B47","doi-asserted-by":"crossref","first-page":"811","DOI":"10.1007\/s40262-015-0270-6","article-title":"A comprehensive review of drug\u2013drug interactions with metformin","volume":"54","author":"Stage","year":"2015","journal-title":"Clin. Pharmacokinet"},{"key":"2023051605023635700_bty294-B48","doi-asserted-by":"crossref","first-page":"8481.","DOI":"10.1038\/ncomms9481","article-title":"Combining genomic and network characteristics for extended capability in predicting synergistic drugs for cancer","volume":"6","author":"Sun","year":"2015","journal-title":"Nat. Commun"},{"key":"2023051605023635700_bty294-B49","doi-asserted-by":"crossref","first-page":"D380","DOI":"10.1093\/nar\/gkv1277","article-title":"STITCH 5: augmenting protein\u2013chemical interaction networks with tissue and affinity data","volume":"44","author":"Szklarczyk","year":"2016","journal-title":"Nucleic Acids Res"},{"key":"2023051605023635700_bty294-B50","doi-asserted-by":"crossref","first-page":"D362","DOI":"10.1093\/nar\/gkw937","article-title":"The STRING database in 2017: quality-controlled protein\u2013protein association networks, made broadly accessible","volume":"45","author":"Szklarczyk","year":"2017","journal-title":"Nucleic Acids Res"},{"key":"2023051605023635700_bty294-B51","doi-asserted-by":"crossref","first-page":"16","DOI":"10.1186\/s13321-017-0200-8","article-title":"Predicting drug\u2013drug interactions through drug structural similarities and interaction networks incorporating pharmacokinetics and pharmacodynamics knowledge","volume":"9","author":"Takeda","year":"2017","journal-title":"J. Cheminform"},{"key":"2023051605023635700_bty294-B52","doi-asserted-by":"crossref","first-page":"125ra31.","DOI":"10.1126\/scitranslmed.3003377","article-title":"Data-driven prediction of drug effects and interactions","volume":"4","author":"Tatonetti","year":"2012","journal-title":"Sci. Transl. Med"},{"key":"2023051605023635700_bty294-B53","first-page":"2071","author":"Trouillon","year":"2016"},{"key":"2023051605023635700_bty294-B54","doi-asserted-by":"crossref","first-page":"1066","DOI":"10.1136\/amiajnl-2012-000935","article-title":"Drug-drug interaction through molecular structure similarity analysis","volume":"19","author":"Vilar","year":"2012","journal-title":"J. Am. Med. Inform. Assoc"},{"key":"2023051605023635700_bty294-B55","author":"Vilar","year":"2017"},{"key":"2023051605023635700_bty294-B56","doi-asserted-by":"crossref","first-page":"12339","DOI":"10.1038\/srep12339","article-title":"Label propagation prediction of drug-drug interactions based on clinical side effects","volume":"5","author":"Zhang","year":"2015","journal-title":"Sci. Rep"},{"key":"2023051605023635700_bty294-B57","doi-asserted-by":"crossref","first-page":"18.","DOI":"10.1186\/s12859-016-1415-9","article-title":"Predicting potential drug-drug interactions by integrating chemical, biological, phenotypic and network data","volume":"18","author":"Zhang","year":"2017","journal-title":"BMC Bioinformatics"},{"key":"2023051605023635700_bty294-B58","doi-asserted-by":"crossref","first-page":"e1002323.","DOI":"10.1371\/journal.pcbi.1002323","article-title":"Prediction of drug combinations by integrating molecular and pharmacological data","volume":"7","author":"Zhao","year":"2011","journal-title":"PLoS Comput. Biol"},{"key":"2023051605023635700_bty294-B59","doi-asserted-by":"crossref","first-page":"41","DOI":"10.1109\/TPAMI.2014.2343973","article-title":"Data fusion by matrix factorization","volume":"37","author":"Zitnik","year":"2015","journal-title":"IEEE Trans. Patt. Anal. Mach. Intell"},{"key":"2023051605023635700_bty294-B60","volume-title":"Pacific Symposium on Biocomputing","author":"Zitnik","year":"2016"},{"key":"2023051605023635700_bty294-B61","doi-asserted-by":"crossref","first-page":"2337","DOI":"10.1093\/bioinformatics\/btx160","article-title":"Deep mining heterogeneous networks of biomedical linked data to predict novel drug\u2013target associations","volume":"33","author":"Zong","year":"2017","journal-title":"Bioinformatics"}],"container-title":["Bioinformatics"],"original-title":[],"language":"en","link":[{"URL":"https:\/\/academic.oup.com\/bioinformatics\/article-pdf\/34\/13\/i457\/50316205\/bioinformatics_34_13_i457.pdf","content-type":"application\/pdf","content-version":"vor","intended-application":"syndication"},{"URL":"https:\/\/academic.oup.com\/bioinformatics\/article-pdf\/34\/13\/i457\/50316205\/bioinformatics_34_13_i457.pdf","content-type":"unspecified","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2023,5,16]],"date-time":"2023-05-16T01:04:12Z","timestamp":1684199052000},"score":1,"resource":{"primary":{"URL":"https:\/\/academic.oup.com\/bioinformatics\/article\/34\/13\/i457\/5045770"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2018,6,27]]},"references-count":61,"journal-issue":{"issue":"13","published-print":{"date-parts":[[2018,7,1]]}},"URL":"https:\/\/doi.org\/10.1093\/bioinformatics\/bty294","relation":{"has-preprint":[{"id-type":"doi","id":"10.1101\/258814","asserted-by":"object"}]},"ISSN":["1367-4803","1367-4811"],"issn-type":[{"value":"1367-4803","type":"print"},{"value":"1367-4811","type":"electronic"}],"subject":[],"published-other":{"date-parts":[[2018,7,1]]},"published":{"date-parts":[[2018,6,27]]}}}