{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2024,6,26]],"date-time":"2024-06-26T12:36:31Z","timestamp":1719405391084},"reference-count":48,"publisher":"Springer Science and Business Media LLC","issue":"1","license":[{"start":{"date-parts":[[2024,1,30]],"date-time":"2024-01-30T00:00:00Z","timestamp":1706572800000},"content-version":"tdm","delay-in-days":0,"URL":"https:\/\/creativecommons.org\/licenses\/by\/4.0"},{"start":{"date-parts":[[2024,1,30]],"date-time":"2024-01-30T00:00:00Z","timestamp":1706572800000},"content-version":"vor","delay-in-days":0,"URL":"https:\/\/creativecommons.org\/licenses\/by\/4.0"}],"content-domain":{"domain":["link.springer.com"],"crossmark-restriction":false},"short-container-title":["BMC Bioinformatics"],"abstract":"Abstract<\/jats:title>Drug\u2013drug interactions (DDI) are a critical concern in healthcare due to their potential to cause adverse effects and compromise patient safety. Supervised machine learning models for DDI prediction need to be optimized to learn abstract, transferable features, and generalize to larger chemical spaces, primarily due to the scarcity of high-quality labeled DDI data. Inspired by recent advances in computer vision, we present SMR\u2013DDI, a self-supervised framework that leverages contrastive learning to embed drugs into a scaffold-based feature space. Molecular scaffolds represent the core structural motifs that drive pharmacological activities, making them valuable for learning informative representations. Specifically, we pre-trained SMR\u2013DDI on a large-scale unlabeled molecular dataset. We generated augmented views for each molecule via SMILES enumeration and optimized the embedding process through contrastive loss minimization between views. This enables the model to capture relevant and robust molecular features while reducing noise. We then transfer the learned representations for the downstream prediction of DDI. Experiments show that the new feature space has comparable expressivity to state-of-the-art molecular representations and achieved competitive DDI prediction results while training on less data. Additional investigations also revealed that pre-training on more extensive and diverse unlabeled molecular datasets improved the model\u2019s capability to embed molecules more effectively. Our results highlight contrastive learning as a promising approach for DDI prediction that can identify potentially hazardous drug combinations using only structural information.<\/jats:p>","DOI":"10.1186\/s12859-024-05643-7","type":"journal-article","created":{"date-parts":[[2024,1,30]],"date-time":"2024-01-30T13:03:04Z","timestamp":1706619784000},"update-policy":"http:\/\/dx.doi.org\/10.1007\/springer_crossmark_policy","source":"Crossref","is-referenced-by-count":1,"title":["Learning self-supervised molecular representations for drug\u2013drug interaction prediction"],"prefix":"10.1186","volume":"25","author":[{"given":"Rogia","family":"Kpanou","sequence":"first","affiliation":[]},{"given":"Patrick","family":"Dallaire","sequence":"additional","affiliation":[]},{"given":"Elsa","family":"Rousseau","sequence":"additional","affiliation":[]},{"given":"Jacques","family":"Corbeil","sequence":"additional","affiliation":[]}],"member":"297","published-online":{"date-parts":[[2024,1,30]]},"reference":[{"key":"5643_CR1","doi-asserted-by":"crossref","unstructured":"Carracedo-Reboredo P, Li\u00f1ares-Blanco J. A review on machine learning approaches and trends in drug discovery. PubMed 2021. https:\/\/pubmed.ncbi.nlm.nih.gov\/34471498\/","DOI":"10.1016\/j.csbj.2021.08.011"},{"key":"5643_CR2","doi-asserted-by":"crossref","unstructured":"Ryu JY, & Kim HU. Deep learning improves prediction of drug-drug and drug-food interactions. PubMed 2018. https:\/\/pubmed.ncbi.nlm.nih.gov\/29666228\/","DOI":"10.1073\/pnas.1803294115"},{"key":"5643_CR3","doi-asserted-by":"crossref","unstructured":"Vo TH, Nguyen NTK. Improved prediction of drug-drug interactions using ensemble deep neural networks. Med Drug Discov 2023. https:\/\/hub.tmu.edu.tw\/en\/publications\/improved-prediction-of-drug-drug-interactions-using-ensemble-deep","DOI":"10.1016\/j.medidd.2022.100149"},{"key":"5643_CR4","doi-asserted-by":"crossref","unstructured":"Vo TH, Kim Nguyen NT, Kha QH, Khanh Le NQ. On the road to explainable AI in drug-drug interactions prediction: a systematic review. PubMed 2022. https:\/\/pubmed.ncbi.nlm.nih.gov\/35832629\/","DOI":"10.1016\/j.csbj.2022.04.021"},{"key":"5643_CR5","doi-asserted-by":"crossref","unstructured":"Rohani N, Eslahchi C. Drug-drug interaction predicting by neural network using integrated similarity. PubMed 2019. https:\/\/pubmed.ncbi.nlm.nih.gov\/31541145\/","DOI":"10.1038\/s41598-019-50121-3"},{"key":"5643_CR6","doi-asserted-by":"crossref","unstructured":"Guo L, Lei X. MSResG: using GAE and residual GCN to predict drug-drug interactions based on multi-source drug features. PubMed 2023. https:\/\/pubmed.ncbi.nlm.nih.gov\/36646843\/","DOI":"10.1007\/s12539-023-00550-6"},{"key":"5643_CR7","doi-asserted-by":"crossref","unstructured":"Huang K. [2004.14949] SkipGNN: predicting molecular interactions with skip-graph networks. arXiv 2020. https:\/\/arxiv.org\/abs\/2004.14949","DOI":"10.1038\/s41598-020-77766-9"},{"key":"5643_CR8","doi-asserted-by":"crossref","unstructured":"Al-Rabeah MH, Lakizadeh A. Prediction of drug-drug interaction events using graph neural networks based feature extraction. PubMed 2022. https:\/\/pubmed.ncbi.nlm.nih.gov\/36114278\/","DOI":"10.1038\/s41598-022-19999-4"},{"key":"5643_CR9","doi-asserted-by":"publisher","DOI":"10.1186\/s12859-020-03724-x","author":"Y Feng","year":"2020","unstructured":"Feng Y, Shi Y. DPDDI: a deep predictor for drug-drug interactions - BMC Bioinformatics. BMC Bioinform. 2020. https:\/\/doi.org\/10.1186\/s12859-020-03724-x.","journal-title":"BMC Bioinform"},{"key":"5643_CR10","doi-asserted-by":"crossref","unstructured":"Mei S, Zhang K. A machine learning framework for predicting drug-drug interactions. PubMed 2021. https:\/\/pubmed.ncbi.nlm.nih.gov\/34475500\/","DOI":"10.21203\/rs.3.rs-503867\/v1"},{"key":"5643_CR11","doi-asserted-by":"publisher","DOI":"10.1186\/s12859-023-05242-y","author":"Z Yang","year":"2023","unstructured":"Yang Z, Jin S, Wang S. CNN-Siam: multimodal siamese CNN-based deep learning approach for drug-drug interaction prediction. BMC Bioinform. 2023. https:\/\/doi.org\/10.1186\/s12859-023-05242-y.","journal-title":"BMC Bioinform"},{"key":"5643_CR12","doi-asserted-by":"publisher","DOI":"10.1186\/s12859-022-04612-2","author":"C Zhang","year":"2022","unstructured":"Zhang C, Lu Y. CNN-DDI: a learning-based method for predicting drug\u2013drug interactions using convolution neural networks. BMC Bioinform. 2022. https:\/\/doi.org\/10.1186\/s12859-022-04612-2.","journal-title":"BMC Bioinform"},{"key":"5643_CR13","doi-asserted-by":"publisher","DOI":"10.1007\/s10462-022-10183-8","author":"J Yi-Le Chan","year":"2022","unstructured":"Yi-Le Chan J, Bea KT. State of the art: a review of sentiment analysis based on sequential transfer learning. Artif Intell Rev. 2022. https:\/\/doi.org\/10.1007\/s10462-022-10183-8.","journal-title":"Artif Intell Rev"},{"issue":"6","key":"5643_CR14","doi-asserted-by":"publisher","first-page":"bbab133","DOI":"10.1093\/bib\/bbab133","volume":"22","author":"AK Nyamabo","year":"2021","unstructured":"Nyamabo AK, Yu H, Shi JY. SSI\u2013DDI: substructure\u2013substructure interactions for drug\u2013drug interaction prediction. Brief Bioinform. 2021;22(6):bbab133.","journal-title":"Brief Bioinform"},{"key":"5643_CR15","unstructured":"Deac A, Huang YH, Veli\u010dkovi\u0107 P, Li\u00f2 P, Tang J Drug-drug adverse effect prediction with graph co-attention. arXiv preprint arXiv:1905.00534 (2019)"},{"key":"5643_CR16","doi-asserted-by":"crossref","unstructured":"Feng Y, Zhang S (2022) Prediction of drug-drug interaction using an attention-based graph neural network on drug molecular graphs. MDPI. https:\/\/www.mdpi.com\/1420-3049\/27\/9\/3004","DOI":"10.3390\/molecules27093004"},{"key":"5643_CR17","doi-asserted-by":"publisher","DOI":"10.1186\/s13321-022-00589-5","author":"E Kim","year":"2022","unstructured":"Kim E, Nam H. DeSIDE-DDI: interpretable prediction of drug-drug interactions using drug-induced gene expressions. J Cheminformatics. 2022. https:\/\/doi.org\/10.1186\/s13321-022-00589-5.","journal-title":"J Cheminformatics"},{"key":"5643_CR18","doi-asserted-by":"publisher","DOI":"10.1186\/s12859-021-04398-9","author":"R Kpanou","year":"2021","unstructured":"Kpanou R, Osseni M. On the robustness of generalization of drug\u2013drug interaction models. BMC Bioinform. 2021. https:\/\/doi.org\/10.1186\/s12859-021-04398-9.","journal-title":"BMC Bioinform"},{"key":"5643_CR19","doi-asserted-by":"crossref","unstructured":"Su X, Hu L. Attention-based knowledge graph representation learning for predicting drug-drug interactions. PubMed 2022. https:\/\/pubmed.ncbi.nlm.nih.gov\/35453147\/","DOI":"10.1093\/bib\/bbac140"},{"key":"5643_CR20","doi-asserted-by":"publisher","DOI":"10.1186\/s40537-022-00652-w#Sec5","author":"A Hosna","year":"2022","unstructured":"Hosna A, Merry E. Transfer learning: a friendly introduction. J Big Data. 2022. https:\/\/doi.org\/10.1186\/s40537-022-00652-w#Sec5.","journal-title":"J Big Data"},{"key":"5643_CR21","unstructured":"Zhuang F, Qi Z. [1911.02685] A comprehensive survey on transfer learning. arXiv 2019. https:\/\/arxiv.org\/abs\/1911.02685"},{"key":"5643_CR22","doi-asserted-by":"crossref","unstructured":"Qasim R, Bangyal WH. A fine-tuned BERT-based transfer learning approach for text classification. Hindawi 2022. https:\/\/www.hindawi.com\/journals\/jhe\/2022\/3498123\/","DOI":"10.1155\/2022\/3498123"},{"key":"5643_CR23","doi-asserted-by":"crossref","unstructured":"Kim HE, & Cosa-Linan A. Transfer learning for medical image classification: a literature review. PubMed 2022. https:\/\/pubmed.ncbi.nlm.nih.gov\/35418051\/","DOI":"10.21203\/rs.3.rs-844222\/v1"},{"key":"5643_CR24","unstructured":"Cai C, & Wang S. Transfer learning for drug discovery. PubMed 2020. https:\/\/pubmed.ncbi.nlm.nih.gov\/32672961\/"},{"key":"5643_CR25","unstructured":"Rani V, Nabi ST. Self-supervised learning: a succinct review. PubMed 2023. https:\/\/pubmed.ncbi.nlm.nih.gov\/36713767\/"},{"key":"5643_CR26","unstructured":"Chen T, Kornblith S. A simple framework for contrastive learning of visual representations. arXiv 2020. https:\/\/arxiv.org\/pdf\/2002.05709.pdf. Accessed 23 May 2023."},{"key":"5643_CR27","unstructured":"Caron M, Misra I. [2006.09882] Unsupervised learning of visual features by contrasting cluster assignments. arXiv 2020. https:\/\/arxiv.org\/abs\/2006.09882"},{"issue":"8","key":"5643_CR28","doi-asserted-by":"publisher","first-page":"1742","DOI":"10.1021\/ci200179y","volume":"51","author":"Y Hu","year":"2011","unstructured":"Hu Y, Stumpfe D, Bajorath J. Lessons learned from molecular scaffold analysis. J Chem Inf Model. 2011;51(8):1742\u201353. https:\/\/doi.org\/10.1021\/ci200179y.","journal-title":"J Chem Inf Model"},{"key":"5643_CR29","unstructured":"Bjerrum EJ. SMILES enumeration as data augmentation for neural network modeling of molecules. arXiv preprint arXiv:1703.07076 2017."},{"key":"5643_CR30","unstructured":"DeepChem. https:\/\/github.com\/deepchem\/deepchem"},{"key":"5643_CR31","unstructured":"RDKit: Open-source cheminformatics. https:\/\/www.rdkit.org"},{"key":"5643_CR32","unstructured":"Oord AVD, Li Y, Vinyals O. Representation learning with contrastive predictive coding. arXiv preprint arXiv:1807.03748 2018."},{"issue":"11","key":"5643_CR33","doi-asserted-by":"publisher","first-page":"2884","DOI":"10.1021\/ci300261r","volume":"52","author":"R Todeschini","year":"2012","unstructured":"Todeschini R, Consonni V, Xiang H, Holliday J, Buscema P, Willett P. Similarity coefficients for binary chemoinformatics data: overview and extended comparison using simulated and real data sets. J Chem Inf Model. 2012;52(11):2884\u2013901. https:\/\/doi.org\/10.1021\/ci300261r.","journal-title":"J Chem Inf Model"},{"key":"5643_CR34","doi-asserted-by":"publisher","DOI":"10.1186\/s13321-015-0069-3","author":"D Bajusz","year":"2015","unstructured":"Bajusz D. Why is Tanimoto index an appropriate choice for fingerprint-based similarity calculations? J Cheminformatics. 2015. https:\/\/doi.org\/10.1186\/s13321-015-0069-3.","journal-title":"J Cheminformatics"},{"key":"5643_CR35","doi-asserted-by":"crossref","unstructured":"Bender A, Glen RC. Molecular similarity: a key technique in molecular informatics. RSC Publishing; 2004. https:\/\/pubs.rsc.org\/en\/content\/articlelanding\/2004\/ob\/b409813g","DOI":"10.1039\/b409813g"},{"key":"5643_CR36","doi-asserted-by":"crossref","unstructured":"Willett P. Similarity-based virtual screening using 2D fingerprints. PubMed 2006. https:\/\/pubmed.ncbi.nlm.nih.gov\/17129822\/","DOI":"10.1016\/j.drudis.2006.10.005"},{"key":"5643_CR37","unstructured":"Willett P. Effectiveness of 2D fingerprints for scaffold hopping. PubMed 2011. https:\/\/pubmed.ncbi.nlm.nih.gov\/21452977\/"},{"key":"5643_CR38","first-page":"623","volume":"66","author":"C Gon\u00e7alveseS\u00e1","year":"2011","unstructured":"Gon\u00e7alveseS\u00e1 C, Aa D, Jp D, Th M, Cm F, Gb S, Rm D. Sedative, anxiolytic and antidepressant activities of Citrus limon (Burn) essential oil in mice. Pharmazie. 2011;66:623.","journal-title":"Pharmazie"},{"issue":"2","key":"5643_CR39","first-page":"526","volume":"226","author":"P Soubrie\u0301","year":"1983","unstructured":"Soubrie\u0301 P, Blas C, Ferron A, Glowinski J. Chlordiazepoxide reduces in vivo serotonin release in the basal ganglia of enc\u00e9phale isol\u00e9 but not anesthetized cats: evidence for a dorsal raphe site of action. J Pharmacol Exp Ther. 1983;226(2):526\u201332.","journal-title":"J Pharmacol Exp Ther"},{"key":"5643_CR40","unstructured":"Hahn M. Extended-connectivity fingerprints. PubMed 2010. https:\/\/pubmed.ncbi.nlm.nih.gov\/20426451\/"},{"key":"5643_CR41","unstructured":"Nourse JG. Reoptimization of MDL keys for use in drug discovery. PubMed 2002. https:\/\/pubmed.ncbi.nlm.nih.gov\/12444722\/"},{"key":"5643_CR42","unstructured":"Frey N, Soklaski R, Axelrod S. Neural Scaling of deep chemical models | theoretical and computational chemistry. ChemRxiv 2022. https:\/\/chemrxiv.org\/engage\/chemrxiv\/article-details\/627bddd544bdd532395fb4b5"},{"key":"5643_CR43","unstructured":"Ahmad W, Simon E. [2209.01712] ChemBERTa-2: towards chemical foundation models. arXiv 2022. https:\/\/arxiv.org\/abs\/2209.01712"},{"key":"5643_CR44","unstructured":"Hu W, Liu B. [1905.12265] Strategies for pre-training graph neural networks. arXiv 2019. https:\/\/arxiv.org\/abs\/1905.12265"},{"key":"5643_CR45","doi-asserted-by":"crossref","unstructured":"Jaeger, S., Fulle, S., & Turk1, S. Mol2vec: unsupervised machine learning approach with chemical intuition. PubMed 2018. https:\/\/pubmed.ncbi.nlm.nih.gov\/29268609\/","DOI":"10.26434\/chemrxiv.5513581.v1"},{"key":"5643_CR46","doi-asserted-by":"publisher","first-page":"52","DOI":"10.1177\/14738716221130338","volume":"22","author":"H Li","year":"2022","unstructured":"Li H, Wang J, Zheng Y, Wang L, Zhang W, Shen H. Compressing and interpreting word embeddings with latent space regularization and interactive semantics probing. Inf Vis. 2022;22:52\u201368. https:\/\/doi.org\/10.1177\/14738716221130338.","journal-title":"Inf Vis"},{"key":"5643_CR47","doi-asserted-by":"publisher","first-page":"99","DOI":"10.1007\/s11634-020-00386-8","volume":"15","author":"L Labiod","year":"2020","unstructured":"Labiod L, Nadif M. Efficient regularized spectral data embedding. Adv Data Anal Classif. 2020;15:99\u2013119. https:\/\/doi.org\/10.1007\/s11634-020-00386-8.","journal-title":"Adv Data Anal Classif"},{"key":"5643_CR48","doi-asserted-by":"publisher","DOI":"10.3389\/fphar.2020.565644\/full","author":"D Polykovskiy","year":"2020","unstructured":"Polykovskiy D. Molecular sets (MOSES): a benchmarking platform for molecular generation models. Frontiers. 2020. https:\/\/doi.org\/10.3389\/fphar.2020.565644\/full.","journal-title":"Frontiers"}],"container-title":["BMC Bioinformatics"],"original-title":[],"language":"en","link":[{"URL":"https:\/\/link.springer.com\/content\/pdf\/10.1186\/s12859-024-05643-7.pdf","content-type":"application\/pdf","content-version":"vor","intended-application":"text-mining"},{"URL":"https:\/\/link.springer.com\/article\/10.1186\/s12859-024-05643-7\/fulltext.html","content-type":"text\/html","content-version":"vor","intended-application":"text-mining"},{"URL":"https:\/\/link.springer.com\/content\/pdf\/10.1186\/s12859-024-05643-7.pdf","content-type":"application\/pdf","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2024,1,30]],"date-time":"2024-01-30T13:03:18Z","timestamp":1706619798000},"score":1,"resource":{"primary":{"URL":"https:\/\/bmcbioinformatics.biomedcentral.com\/articles\/10.1186\/s12859-024-05643-7"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2024,1,30]]},"references-count":48,"journal-issue":{"issue":"1","published-online":{"date-parts":[[2024,12]]}},"alternative-id":["5643"],"URL":"https:\/\/doi.org\/10.1186\/s12859-024-05643-7","relation":{},"ISSN":["1471-2105"],"issn-type":[{"value":"1471-2105","type":"electronic"}],"subject":[],"published":{"date-parts":[[2024,1,30]]},"assertion":[{"value":"20 September 2023","order":1,"name":"received","label":"Received","group":{"name":"ArticleHistory","label":"Article History"}},{"value":"5 January 2024","order":2,"name":"accepted","label":"Accepted","group":{"name":"ArticleHistory","label":"Article History"}},{"value":"30 January 2024","order":3,"name":"first_online","label":"First Online","group":{"name":"ArticleHistory","label":"Article History"}},{"order":1,"name":"Ethics","group":{"name":"EthicsHeading","label":"Declarations"}},{"value":"Not applicable.","order":2,"name":"Ethics","group":{"name":"EthicsHeading","label":"Ethical approval and consent to participate"}},{"value":"Not applicable.","order":3,"name":"Ethics","group":{"name":"EthicsHeading","label":"Consent for publication"}},{"value":"Not applicable.","order":4,"name":"Ethics","group":{"name":"EthicsHeading","label":"Competing interests"}}],"article-number":"47"}}