{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2026,3,22]],"date-time":"2026-03-22T13:14:26Z","timestamp":1774185266214,"version":"3.50.1"},"reference-count":62,"publisher":"Oxford University Press (OUP)","issue":"6","license":[{"start":{"date-parts":[[2025,12,31]],"date-time":"2025-12-31T00:00:00Z","timestamp":1767139200000},"content-version":"vor","delay-in-days":60,"URL":"https:\/\/creativecommons.org\/licenses\/by\/4.0\/"}],"funder":[{"DOI":"10.13039\/100000001","name":"National Science Foundation","doi-asserted-by":"publisher","award":["DBI-2233967"],"award-info":[{"award-number":["DBI-2233967"]}],"id":[{"id":"10.13039\/100000001","id-type":"DOI","asserted-by":"publisher"}]},{"name":"College of Engineering at Virginia Tech"}],"content-domain":{"domain":[],"crossmark-restriction":false},"short-container-title":[],"published-print":{"date-parts":[[2025,11,1]]},"abstract":"<jats:title>Abstract<\/jats:title>\n                  <jats:p>Synergistic drug combinations are often used to treat cancer. Experimental exploration of all possibilities is expensive. Deep learning (DL) offers a potential alternative for predicting drug pair synergy in specific cell lines. However, current methods often suffer from data leakage and lack systematic ablation studies. We propose SynVerse, a comprehensive evaluation framework featuring four data-splitting strategies to assess DL model generalizability and three ablation studies: module-based, feature shuffling, and a novel network-based approach to disentangle factors influencing performance. We evaluated sixteen models incorporating eight drug- and cell line-specific features, five preprocessing techniques, and two encoders. Our analysis revealed that no model outperformed a baseline using one-hot encoding. Biologically meaningful drug or cell line features and drug-drug interactions did not drive predictive performance. All models showed poor generalization to unseen drugs and cell lines. SynVerse highlights the need for substantial improvements before computational predictors can reliably support experimental and clinical settings.<\/jats:p>","DOI":"10.1093\/bib\/bbaf676","type":"journal-article","created":{"date-parts":[[2025,12,4]],"date-time":"2025-12-04T13:09:55Z","timestamp":1764853795000},"source":"Crossref","is-referenced-by-count":1,"title":["SynVerse: a modular framework for building and evaluating deep learning-based drug synergy prediction models"],"prefix":"10.1093","volume":"26","author":[{"ORCID":"https:\/\/orcid.org\/0009-0000-2971-0987","authenticated-orcid":false,"given":"Nure","family":"Tasnina","sequence":"first","affiliation":[{"name":"Department of Computer Science, Virginia Tech , Blacksburg, VA 24060 ,","place":["United States"]}],"role":[{"role":"author","vocabulary":"crossref"}]},{"ORCID":"https:\/\/orcid.org\/0009-0006-3377-0417","authenticated-orcid":false,"given":"Maryam","family":"Haghani","sequence":"additional","affiliation":[{"name":"Department of Computer Science, Virginia Tech , Blacksburg, VA 24060 ,","place":["United States"]}],"role":[{"role":"author","vocabulary":"crossref"}]},{"given":"T M","family":"Murali","sequence":"additional","affiliation":[{"name":"Department of Computer Science, Virginia Tech , Blacksburg, VA 24060 ,","place":["United States"]}],"role":[{"role":"author","vocabulary":"crossref"}]}],"member":"286","published-online":{"date-parts":[[2025,12,31]]},"reference":[{"key":"2025123023164388900_ref1","doi-asserted-by":"publisher","first-page":"166","DOI":"10.1038\/s41586-022-04437-2","article-title":"Effective drug combinations in breast, colon, and pancreatic cancer cells","volume":"603","author":"Jaaks","year":"2022","journal-title":"Nature"},{"key":"2025123023164388900_ref2","doi-asserted-by":"publisher","first-page":"1155","DOI":"10.1158\/1535-7163.MCT-15-0843","article-title":"An unbiased oncology compound screen to identify novel combination strategies","volume":"15","author":"O\u2019Neil","year":"2016","journal-title":"Mol Cancer Ther"},{"key":"2025123023164388900_ref3","doi-asserted-by":"publisher","first-page":"213","DOI":"10.1038\/s41573-022-00615-z","article-title":"Rational combinations of targeted cancer therapies: background, advances and challenges","volume":"22","author":"Jin","year":"2023","journal-title":"Nat Rev Drug Discov"},{"key":"2025123023164388900_ref4","doi-asserted-by":"publisher","first-page":"445","DOI":"10.1038\/nrclinonc.2015.61","article-title":"Targeting notch, hedgehog, and wnt pathways in cancer stem cells: clinical update","volume":"12","author":"Takebe","year":"2015","journal-title":"Nat Rev Clin Oncol"},{"key":"2025123023164388900_ref5","doi-asserted-by":"publisher","first-page":"2305","DOI":"10.1158\/0008-5472.CAN-14-3013","article-title":"Pancreatic cancer combination therapy using a BH3 mimetic and a synthetic tetracycline","volume":"75","author":"Quinn","year":"2015","journal-title":"Cancer Res"},{"key":"2025123023164388900_ref6","first-page":"1","article-title":"Synergy, additivism and antagonism in immunosuppression. A critical review","volume":"28","author":"Berenbaum","year":"1977","journal-title":"Clin Exp Immunol"},{"key":"2025123023164388900_ref7","doi-asserted-by":"publisher","first-page":"93","DOI":"10.1016\/S0031-6997(25)00026-2","article-title":"What is synergy?","volume":"41","author":"Berenbaum","year":"1989","journal-title":"Pharmacol Rev"},{"key":"2025123023164388900_ref8","doi-asserted-by":"publisher","first-page":"btae134","DOI":"10.1093\/bioinformatics\/btae134","article-title":"From cell lines to cancer patients: personalized drug synergy prediction","volume":"40","author":"Kuru","year":"2024","journal-title":"Bioinformatics"},{"key":"2025123023164388900_ref9","doi-asserted-by":"publisher","first-page":"57","DOI":"10.1038\/nrclinonc.2016.96","article-title":"Combine and conquer: challenges for targeted therapy combinations in early phase trials","volume":"14","author":"Lopez","year":"2017","journal-title":"Nat Rev Clin Oncol"},{"key":"2025123023164388900_ref10","doi-asserted-by":"publisher","first-page":"W43","DOI":"10.1093\/nar\/gkz337","article-title":"DrugComb: an integrative cancer drug combination data portal","volume":"47","author":"Zagidullin","year":"2019","journal-title":"Nucleic Acids Res"},{"key":"2025123023164388900_ref11","doi-asserted-by":"publisher","first-page":"D871","DOI":"10.1093\/nar\/gkz1007","article-title":"DrugCombDB: a comprehensive database of drug combinations toward the discovery of combinatorial therapy","volume":"48","author":"Liu","year":"2020","journal-title":"Nucleic Acids Res"},{"key":"2025123023164388900_ref12","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":"2025123023164388900_ref13","doi-asserted-by":"publisher","first-page":"e1006752","DOI":"10.1371\/journal.pcbi.1006752","article-title":"Drug combination sensitivity scoring facilitates the discovery of synergistic and efficacious drug combinations in cancer","volume":"15","author":"Malyutina","year":"2019","journal-title":"PLoS Comput Biol"},{"key":"2025123023164388900_ref14","doi-asserted-by":"publisher","first-page":"1538","DOI":"10.1093\/bioinformatics\/btx806","article-title":"DeepSynergy: predicting anti-cancer drug synergy with deep learning","volume":"34","author":"Preuer","year":"2018","journal-title":"Bioinformatics"},{"key":"2025123023164388900_ref15","article-title":"MatchMaker: a deep learning framework for drug synergy prediction","volume":"19","author":"Brahim","year":"2021","journal-title":"IEEE\/ACM Trans Comput Biol Bioinform"},{"key":"2025123023164388900_ref16","doi-asserted-by":"publisher","first-page":"btad177","DOI":"10.1093\/bioinformatics\/btad177","article-title":"MARSY: a multitask deep-learning framework for prediction of drug combination synergy scores","volume":"39","author":"El Khili","year":"2023","journal-title":"Bioinformatics"},{"key":"2025123023164388900_ref17","doi-asserted-by":"publisher","first-page":"223","DOI":"10.1007\/978-1-0716-0849-4_12","article-title":"Synergistic drug combination prediction by integrating multiomics data in deep learning models","volume-title":"Translational Bioinformatics for Therapeutic Development","author":"Zhang","year":"2021"},{"key":"2025123023164388900_ref18","doi-asserted-by":"publisher","first-page":"3689","DOI":"10.1021\/acs.jcim.4c00165","article-title":"MMSyn: a new multimodal deep learning framework for enhanced prediction of synergistic drug combinations","volume":"64","author":"Pang","year":"2024","journal-title":"J Chem Inf Model"},{"key":"2025123023164388900_ref19","doi-asserted-by":"publisher","first-page":"bbab390","DOI":"10.1093\/bib\/bbab390","article-title":"DeepDDS: deep graph neural network with attention mechanism to predict synergistic drug combinations","volume":"23","author":"Wang","year":"2022","journal-title":"Brief Bioinform"},{"key":"2025123023164388900_ref20","doi-asserted-by":"publisher","first-page":"e1008653","DOI":"10.1371\/journal.pcbi.1008653","article-title":"TranSynergy: mechanism-driven interpretable deep neural network for the synergistic prediction and pathway deconvolution of drug combinations","volume":"17","author":"Liu","year":"2021","journal-title":"PLoS Comput Biol"},{"key":"2025123023164388900_ref21","doi-asserted-by":"publisher","first-page":"bbac302","DOI":"10.1093\/bib\/bbac302","article-title":"DTSyn: a dual-transformer-based neural network to predict synergistic drug combinations","volume":"23","author":"Hu","year":"2022","journal-title":"Brief Bioinform"},{"key":"2025123023164388900_ref22","doi-asserted-by":"publisher","first-page":"33","DOI":"10.1186\/s13321-023-00690-3","article-title":"DFFNDDS: prediction of synergistic drug combinations with dual feature fusion networks","volume":"15","author":"Xu","year":"2023","journal-title":"J Chem"},{"key":"2025123023164388900_ref23","doi-asserted-by":"publisher","first-page":"bbae015","DOI":"10.1093\/bib\/bbae015","article-title":"SynergyX: a multi-modality mutual attention network for interpretable drug synergy prediction","volume":"25","author":"Guo","year":"2024","journal-title":"Brief Bioinform"},{"key":"2025123023164388900_ref24","doi-asserted-by":"publisher","first-page":"42","DOI":"10.1093\/jamia\/ocaa212","article-title":"Anticancer drug synergy prediction in understudied tissues using transfer learning","volume":"28","author":"Kim","year":"2021","journal-title":"J Am Med Inform Assoc"},{"key":"2025123023164388900_ref25","doi-asserted-by":"publisher","first-page":"762","DOI":"10.1158\/1541-7786.MCR-21-0735","article-title":"SynPathy: predicting drug synergy through drug-associated pathways using deep learning","volume":"20","author":"Tang","year":"2022","journal-title":"Mol Cancer Res"},{"key":"2025123023164388900_ref26","doi-asserted-by":"publisher","first-page":"316","DOI":"10.1007\/s12539-023-00558-y","article-title":"Predicting drug synergy and discovering new drug combinations based on a graph autoencoder and convolutional neural network","volume":"15","author":"Li","year":"2023","journal-title":"Interdiscip Sci Comput Life Sci"},{"key":"2025123023164388900_ref27","doi-asserted-by":"publisher","first-page":"6184","DOI":"10.1038\/s41598-023-33271-3","article-title":"SYNDEEP: a deep learning approach for the prediction of cancer drugs synergy","volume":"13","author":"Torkamannia","year":"2023","journal-title":"Sci Rep"},{"key":"2025123023164388900_ref28","doi-asserted-by":"publisher","first-page":"btad438","DOI":"10.1093\/bioinformatics\/btad438","article-title":"DeepTraSynergy: drug combinations using multimodal deep learning with transformers","volume":"39","author":"Rafiei","year":"2023","journal-title":"Bioinformatics"},{"key":"2025123023164388900_ref29","doi-asserted-by":"publisher","first-page":"e1010200","DOI":"10.1371\/journal.pcbi.1010200","article-title":"A systematic evaluation of deep learning methods for the prediction of drug synergy in cancer","volume":"19","author":"Baptista","year":"2023","journal-title":"PLoS Comput Biol"},{"key":"2025123023164388900_ref30","doi-asserted-by":"publisher","first-page":"1134","DOI":"10.1038\/nmeth.2259","article-title":"Flaws in evaluation schemes for pair-input computational predictions","volume":"9","author":"Park","year":"2012","journal-title":"Nat Methods"},{"key":"2025123023164388900_ref31","doi-asserted-by":"publisher","first-page":"1437","DOI":"10.1016\/j.cell.2017.10.049","article-title":"A next generation connectivity map: L1000 platform and the first 1,000,000 profiles","volume":"171","author":"Subramanian","year":"2017","journal-title":"Cell"},{"key":"2025123023164388900_ref32","doi-asserted-by":"publisher","first-page":"9431","DOI":"10.1038\/s41467-024-53751-y","article-title":"MolE: a foundation model for molecular graphs using disentangled attention","volume":"15","author":"M\u00e9ndez-Lucio","year":"2024","journal-title":"Nat Commun"},{"key":"2025123023164388900_ref33","doi-asserted-by":"publisher","first-page":"7568","DOI":"10.1038\/s41467-023-43214-1","article-title":"A knowledge-guided pre-training framework for improving molecular representation learning","volume":"14","author":"Li","year":"2023","journal-title":"Nat Commun"},{"key":"2025123023164388900_ref34","doi-asserted-by":"publisher","first-page":"2323","DOI":"10.1038\/s41467-024-46440-3","article-title":"Bidirectional generation of structure and properties through a single molecular foundation model","volume":"15","author":"Chang","year":"2024","journal-title":"Nat Commun"},{"key":"2025123023164388900_ref35","article-title":"Semi-supervised classification with graph convolutional networks","volume-title":"International Conference on Learning Representations (ICLR)","author":"Kipf"},{"key":"2025123023164388900_ref36","doi-asserted-by":"crossref","first-page":"785","DOI":"10.1145\/2939672.2939785","article-title":"Xgboost: a scalable tree boosting system","volume-title":"Proceedings of the 22nd ACM SIGKDD International Conference on Knowledge Discovery and Data mining","author":"Chen","year":"2016"},{"key":"2025123023164388900_ref37","doi-asserted-by":"publisher","first-page":"5","DOI":"10.1023\/A:1010933404324","article-title":"Random forests","volume":"45","author":"Breiman","year":"2001","journal-title":"Mach Learn"},{"key":"2025123023164388900_ref38","doi-asserted-by":"crossref","DOI":"10.58647\/REXPO.25000096.v1","article-title":"From hype to health check: critical evaluation of drug response prediction models with dreval","volume-title":"RExPO25 Conference","author":"List","year":"2025"},{"key":"2025123023164388900_ref39","article-title":"Leak proof PDBBind: a reorganized dataset of protein-ligand complexes for more generalizable binding affinity prediction","author":"Li"},{"key":"2025123023164388900_ref40","doi-asserted-by":"publisher","first-page":"bbae076","DOI":"10.1093\/bib\/bbae076","article-title":"Cracking the black box of deep sequence-based protein-protein interaction prediction","volume":"25","author":"Bernett","year":"2024","journal-title":"Brief Bioinform"},{"key":"2025123023164388900_ref41","doi-asserted-by":"publisher","first-page":"2025","DOI":"10.1101\/2025.01.23.634511","article-title":"Benchmarking data leakage on link prediction in biomedical knowledge graph embeddings","author":"Briere","year":"2025"},{"key":"2025123023164388900_ref42","doi-asserted-by":"publisher","first-page":"1989","DOI":"10.1038\/s41467-023-37572-z","article-title":"Improving the generalizability of protein-ligand binding predictions with AI-bind","volume":"14","author":"Chatterjee","year":"2023","journal-title":"Nat Commun"},{"key":"2025123023164388900_ref43","doi-asserted-by":"publisher","first-page":"665","DOI":"10.1038\/s42256-020-00257-z","article-title":"Shortcut learning in deep neural networks","volume":"2","author":"Geirhos","year":"2020","journal-title":"Nat Mach Intell"},{"key":"2025123023164388900_ref44","doi-asserted-by":"publisher","first-page":"3337","DOI":"10.1038\/s41467-025-58606-8","article-title":"Data splitting to avoid information leakage with datasail","volume":"16","author":"Joeres","year":"2025","journal-title":"Nat Commun"},{"key":"2025123023164388900_ref45","doi-asserted-by":"publisher","first-page":"btad351","DOI":"10.1093\/bioinformatics\/btad351","article-title":"A complete graph-based approach with multi-task learning for predicting synergistic drug combinations","volume":"39","author":"Wang","year":"2023","journal-title":"Bioinformatics"},{"key":"2025123023164388900_ref46","doi-asserted-by":"publisher","first-page":"31","DOI":"10.1021\/ci00057a005","article-title":"SMILES, a chemical language and information system. 1. Introduction to methodology and encoding rules","volume":"28","author":"Weininger","year":"1988","journal-title":"J Chem Inf Comput Sci"},{"key":"2025123023164388900_ref47","volume-title":"PubChemPy Documentation","author":"Swain","year":"2014"},{"key":"2025123023164388900_ref48","doi-asserted-by":"publisher","first-page":"W563","DOI":"10.1093\/nar\/gky294","article-title":"An update on PUG-REST: RESTful interface for programmatic access to PubChem","volume":"46","author":"Kim","year":"2018","journal-title":"Nucleic Acids Res"},{"key":"2025123023164388900_ref49","doi-asserted-by":"publisher","first-page":"1273","DOI":"10.1021\/ci010132r","article-title":"Reoptimization of MDL keys for use in drug discovery","volume":"42","author":"Durant","year":"2002","journal-title":"J Chem Inf Comput Sci"},{"key":"2025123023164388900_ref50","doi-asserted-by":"publisher","first-page":"107","DOI":"10.1021\/c160017a018","article-title":"The generation of a unique machine description for chemical structures-a technique developed at chemical abstracts service","volume":"5","author":"Morgan","year":"1965","journal-title":"J Chem Doc"},{"key":"2025123023164388900_ref51","doi-asserted-by":"publisher","first-page":"742","DOI":"10.1021\/ci100050t","article-title":"Extended-connectivity fingerprints","volume":"50","author":"Rogers","year":"2010","journal-title":"J Chem Inf Model"},{"key":"2025123023164388900_ref52","volume-title":"RDKit: Open-Source Cheminformatics","author":"Landrum","year":"2006"},{"key":"2025123023164388900_ref53","doi-asserted-by":"publisher","first-page":"D1465","DOI":"10.1093\/nar\/gkad751","article-title":"TTD: therapeutic target database describing target druggability information","volume":"52","author":"Zhou","year":"2024","journal-title":"Nucleic Acids Res"},{"key":"2025123023164388900_ref54","volume-title":"Deep Learning for the Life Sciences","author":"Ramsundar","year":"2019"},{"key":"2025123023164388900_ref55","doi-asserted-by":"publisher","first-page":"603","DOI":"10.1038\/nature11003","article-title":"The cancer cell line Encyclopedia enables predictive modelling of anticancer drug sensitivity","volume":"483","author":"Barretina","year":"2012","journal-title":"Nature"},{"key":"2025123023164388900_ref56","doi-asserted-by":"publisher","first-page":"588","DOI":"10.1002\/psp4.12107","article-title":"Systematic quality control analysis of LINCS data","volume":"5","author":"Cheng","year":"2016","journal-title":"CPT Pharmacometrics Syst Pharmacol"},{"key":"2025123023164388900_ref57","doi-asserted-by":"publisher","first-page":"D638","DOI":"10.1093\/nar\/gkac1000","article-title":"The STRING database in 2023: protein\u2013protein association networks and functional enrichment analyses for any sequenced genome of interest","volume":"51","author":"Szklarczyk","year":"2023","journal-title":"Nucleic Acids Res"},{"key":"2025123023164388900_ref58","first-page":"23","article-title":"A new algorithm for data compression","volume":"12","author":"Gage","year":"1994","journal-title":"The C Users Journal"},{"key":"2025123023164388900_ref59","article-title":"Attention is all you need","volume":"30","author":"Vaswani","year":"2017","journal-title":"Advances in Neural Information Processing Systems"},{"key":"2025123023164388900_ref60","doi-asserted-by":"publisher","first-page":"e39038","DOI":"10.1016\/j.heliyon.2024.e39038","article-title":"Transformer-based models for chemical SMILES representation: a comprehensive literature review","volume":"10","author":"Mswahili","year":"2024","journal-title":"Heliyon"},{"key":"2025123023164388900_ref61","doi-asserted-by":"publisher","first-page":"910","DOI":"10.1126\/science.1065103","article-title":"Specificity and stability in topology of protein networks","volume":"296","author":"Maslov","year":"2002","journal-title":"Science"},{"key":"2025123023164388900_ref62","doi-asserted-by":"publisher","first-page":"48","DOI":"10.1038\/s43588-024-00735-z","article-title":"A simulated annealing algorithm for randomizing weighted networks","volume":"5","author":"Milisav","year":"2025","journal-title":"Nat Comput Sci"}],"container-title":["Briefings in Bioinformatics"],"original-title":[],"language":"en","link":[{"URL":"https:\/\/academic.oup.com\/bib\/article-pdf\/26\/6\/bbaf676\/66168521\/bbaf676.pdf","content-type":"application\/pdf","content-version":"vor","intended-application":"syndication"},{"URL":"https:\/\/academic.oup.com\/bib\/article-pdf\/26\/6\/bbaf676\/66168521\/bbaf676.pdf","content-type":"unspecified","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2025,12,31]],"date-time":"2025-12-31T04:16:51Z","timestamp":1767154611000},"score":1,"resource":{"primary":{"URL":"https:\/\/academic.oup.com\/bib\/article\/doi\/10.1093\/bib\/bbaf676\/8407512"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2025,11,1]]},"references-count":62,"journal-issue":{"issue":"6","published-print":{"date-parts":[[2025,11,1]]}},"URL":"https:\/\/doi.org\/10.1093\/bib\/bbaf676","relation":{},"ISSN":["1467-5463","1477-4054"],"issn-type":[{"value":"1467-5463","type":"print"},{"value":"1477-4054","type":"electronic"}],"subject":[],"published-other":{"date-parts":[[2025,11]]},"published":{"date-parts":[[2025,11,1]]},"article-number":"bbaf676"}}