Predicting Drug-Drug Interactions Using Knowledge Graphs

Predicting Drug-Drug Interactions Using Knowledge Graphs | Research Square window.SnipcartSettings = { analytics: { enabled: false } }; (function() { var accessVector = localStorage.getItem('access_vector') || ''; window.dataLayer = window.dataLayer || []; if (accessVector) { window.dataLayer.push({ user: { profile: { profileInfo: { snid: accessVector } } } }); } })(); (function(w,d,s,l,i){w[l]=w[l]||[];w[l].push({'gtm.start':new Date().getTime(),event:'gtm.js'});var f=d.getElementsByTagName(s)[0],j=d.createElement(s),dl=l!='dataLayer'?'&l='+l:'';j.async=true;j.src='https://www.googletagmanager.com/gtm.js?id='+i+dl;f.parentNode.insertBefore(j,f);})(window,document,'script','dataLayer','GTM-K279D39R'); Browse Preprints In Review Journals COVID-19 Preprints AJE Video Bytes Research Tools Research Promotion AJE Professional Editing AJE Rubriq About Preprint Platform In Review Editorial Policies Our Team Advisory Board Help Center Sign In Submit a Preprint Cite Share Download PDF Research Article Predicting Drug-Drug Interactions Using Knowledge Graphs Lizzy Farrugia, Jeremy Debattista, Lilian M. Azzopardi, Charlie Abela This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-4492557/v1 This work is licensed under a CC BY 4.0 License Status: Posted Version 1 posted You are reading this latest preprint version Abstract In the last decades, people have been consuming and combining more drugsthan before, increasing the number of Drug-Drug Interactions (DDIs). To pre-dict unknown DDIs, recently, studies started incorporating Knowledge Graphs(KGs) since they are able to capture the relationships among entities provid-ing better drug representations than using a single drug property. In this paper,we propose an end-to-end framework that integrates several drug features frompublic drug repositories into a KG and embeds the nodes in the graph using var-ious translation, factorisation and Neural Network (NN) based KG Embedding(KGE) methods. Ultimately, we use a Machine Learning (ML) algorithm thatpredicts unknown DDIs. Among the different translation and factorisation-basedKGE models, we found that the best performing combination was the ComplExembedding method with a Long Short-Term Memory (LSTM) network, whichobtained an F 1-score of 95.19% on a dataset based on the DDIs found in Drug-Bank version 5.1.8. This score is 5.61% better than the state-of-the-art modelDeepDDI. Additionally, we also developed a graph auto-encoder model that usesa Graph Neural Network (GNN), which achieved an F 1-score of 91.94%. Con-sequently, GNNs have demonstrated a stronger ability to mine the underlyingsemantics of the KG than the ComplEx model, and thus using higher dimensionembeddings within the GNN can lead to state-of-the-art performance. Knowledge Graphs Graph Neural Networks Drug-Drug Interactions Deep Learning Full Text Additional Declarations No competing interests reported. Cite Share Download PDF Status: Posted Version 1 posted You are reading this latest preprint version Research Square lets you share your work early, gain feedback from the community, and start making changes to your manuscript prior to peer review in a journal. As a division of Research Square Company, we’re committed to making research communication faster, fairer, and more useful. We do this by developing innovative software and high quality services for the global research community. Our growing team is made up of researchers and industry professionals working together to solve the most critical problems facing scientific publishing. Also discoverable on Platform About Our Team In Review Editorial Policies Advisory Board Help Center Resources Author Services Accessibility API Access RSS feed Manage Cookie Preferences © Research Square 2026 | ISSN 2693-5015 (online) Privacy Policy Terms of Service Do Not Sell My Personal Information {"props":{"pageProps":{"initialData":{"identity":"rs-4492557","acceptedTermsAndConditions":true,"allowDirectSubmit":true,"archivedVersions":[],"articleType":"Research Article","associatedPublications":[],"authors":[{"id":308257980,"identity":"7d869b9d-a174-4971-9a01-f13915e5c7c8","order_by":0,"name":"Lizzy Farrugia","email":"","orcid":"","institution":"Univeristy of Malta","correspondingAuthor":false,"prefix":"","firstName":"Lizzy","middleName":"","lastName":"Farrugia","suffix":""},{"id":308257981,"identity":"4724643c-02a0-4ccf-bf02-df6de4aec079","order_by":1,"name":"Jeremy Debattista","email":"","orcid":"","institution":"Univeristy of Malta","correspondingAuthor":false,"prefix":"","firstName":"Jeremy","middleName":"","lastName":"Debattista","suffix":""},{"id":308257982,"identity":"88bcd31c-2ac9-49a9-a8d0-bc729ff51573","order_by":2,"name":"Lilian M. Azzopardi","email":"","orcid":"","institution":"Univeristy of Malta","correspondingAuthor":false,"prefix":"","firstName":"Lilian","middleName":"M.","lastName":"Azzopardi","suffix":""},{"id":308257983,"identity":"cd035d74-13a6-4162-a28f-b459af49da7f","order_by":3,"name":"Charlie Abela","email":"data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAZAAAAAyAQMAAABI0h/eAAAABlBMVEX///8AAABVwtN+AAAACXBIWXMAAA7EAAAOxAGVKw4bAAAA8klEQVRIiWNgGAWjYDACCSB+YABmMB5gqACLGRDWkgDRwnCA4QxcCx5tYC1QxgHGNiK08M9ufvYhocAGxDhw4Oe8w9EM7M3bJBhz/uC25M4x4xkJBmkgRsLB3m2Hcxt4jpVJMG7DbYuBRIIx0C+HGRhu5Bgc4AVpkcgxI6Al/TNQy38G+Rv5Hw7+nQPUIv+GkJYckC0HGAxu5DAc5m0A2cKDX4vEjZxioJZkHsM7xwwOyxxLz23jSSu2SNxmjFML/4z0zQwf/tjJyd1ufvjwTY11bj/74Y03Pm6Tw6kFBnjgLDYQkUBQwygYBaNgFIwCfAAAB9RUWWr7cV8AAAAASUVORK5CYII=","orcid":"","institution":"Univeristy of Malta","correspondingAuthor":true,"prefix":"","firstName":"Charlie","middleName":"","lastName":"Abela","suffix":""}],"badges":[],"createdAt":"2024-05-28 17:45:46","currentVersionCode":1,"declarations":"","doi":"10.21203/rs.3.rs-4492557/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-4492557/v1","draftVersion":[],"editorialEvents":[],"editorialNote":"","failedWorkflow":false,"files":[{"id":72156256,"identity":"e36a056f-23c4-48ca-a738-518f4cdf3020","added_by":"auto","created_at":"2024-12-23 08:54:50","extension":"pdf","order_by":1,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":675274,"visible":true,"origin":"","legend":"","description":"","filename":"kais2024predictingdrugdruginteractionsusingKGs.pdf","url":"https://assets-eu.researchsquare.com/files/rs-4492557/v1_covered_a43875a3-5422-46ca-89dd-3f1cc99959d6.pdf"}],"financialInterests":"No competing interests reported.","formattedTitle":"Predicting Drug-Drug Interactions Using Knowledge Graphs","fulltext":[],"fulltextSource":"","fullText":"","funders":[],"hasAdminPriorityOnWorkflow":false,"hasManuscriptDocX":false,"hasOptedInToPreprint":true,"hasPassedJournalQc":"","hasAnyPriority":false,"hideJournal":true,"highlight":"","institution":"","isAcceptedByJournal":false,"isAuthorSuppliedPdf":true,"isDeskRejected":"","isHiddenFromSearch":false,"isInQc":false,"isInWorkflow":false,"isPdf":true,"isPdfUpToDate":true,"isWithdrawnOrRetracted":false,"journal":{"display":true,"email":"[email protected]","identity":"researchsquare","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":true,"externalIdentity":"","sideBox":"","snPcode":"","submissionUrl":"/submission","title":"Research Square","twitterHandle":"researchsquare","acdcEnabled":true,"dfaEnabled":false,"editorialSystem":"","reportingPortfolio":"","inReviewEnabled":false,"inReviewRevisionsEnabled":true},"keywords":"Knowledge Graphs, Graph Neural Networks, Drug-Drug Interactions, Deep Learning","lastPublishedDoi":"10.21203/rs.3.rs-4492557/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-4492557/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"In the last decades, people have been consuming and combining more drugsthan before, increasing the number of Drug-Drug Interactions (DDIs). To pre-dict unknown DDIs, recently, studies started incorporating Knowledge Graphs(KGs) since they are able to capture the relationships among entities provid-ing better drug representations than using a single drug property. In this paper,we propose an end-to-end framework that integrates several drug features frompublic drug repositories into a KG and embeds the nodes in the graph using var-ious translation, factorisation and Neural Network (NN) based KG Embedding(KGE) methods. Ultimately, we use a Machine Learning (ML) algorithm thatpredicts unknown DDIs. Among the different translation and factorisation-basedKGE models, we found that the best performing combination was the ComplExembedding method with a Long Short-Term Memory (LSTM) network, whichobtained an F 1-score of 95.19% on a dataset based on the DDIs found in Drug-Bank version 5.1.8. This score is 5.61% better than the state-of-the-art modelDeepDDI. Additionally, we also developed a graph auto-encoder model that usesa Graph Neural Network (GNN), which achieved an F 1-score of 91.94%. Con-sequently, GNNs have demonstrated a stronger ability to mine the underlyingsemantics of the KG than the ComplEx model, and thus using higher dimensionembeddings within the GNN can lead to state-of-the-art performance.","manuscriptTitle":"Predicting Drug-Drug Interactions Using Knowledge Graphs","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2024-06-11 06:32:50","doi":"10.21203/rs.3.rs-4492557/v1","editorialEvents":[{"type":"communityComments","content":0}],"status":"published","journal":{"display":true,"email":"[email protected]","identity":"researchsquare","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":true,"externalIdentity":"","sideBox":"","snPcode":"","submissionUrl":"/submission","title":"Research Square","twitterHandle":"researchsquare","acdcEnabled":true,"dfaEnabled":false,"editorialSystem":"","reportingPortfolio":"","inReviewEnabled":false,"inReviewRevisionsEnabled":true}}],"origin":"","ownerIdentity":"1a9a8412-6fd2-4ea3-bd69-98656f36c4bd","owner":[],"postedDate":"June 11th, 2024","published":true,"recentEditorialEvents":[],"rejectedJournal":[],"revision":"","amendment":"","status":"posted","subjectAreas":[],"tags":[],"updatedAt":"2024-12-23T08:54:29+00:00","versionOfRecord":[],"versionCreatedAt":"2024-06-11 06:32:50","video":"","vorDoi":"","vorDoiUrl":"","workflowStages":[]},"version":"v1","identity":"rs-4492557","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-4492557","identity":"rs-4492557","version":["v1"]},"buildId":"qtupq5eGEP_6zYnWcrvyt","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}