{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2026,10,15]],"date-time":"2026-10-15T15:47:58Z","timestamp":1792079278740,"version":"4.4.0"},"reference-count":40,"publisher":"Elsevier BV","issue":"20","license":[{"start":{"date-parts":[[2026,10,1]],"date-time":"2026-10-01T00:00:00Z","timestamp":1790812800000},"content-version":"tdm","delay-in-days":0,"URL":"https:\/\/www.elsevier.com\/tdm\/userlicense\/1.0\/"},{"start":{"date-parts":[[2026,10,1]],"date-time":"2026-10-01T00:00:00Z","timestamp":1790812800000},"content-version":"tdm","delay-in-days":0,"URL":"https:\/\/www.elsevier.com\/legal\/tdmrep-license"},{"start":{"date-parts":[[2026,10,1]],"date-time":"2026-10-01T00:00:00Z","timestamp":1790812800000},"content-version":"stm-asf","delay-in-days":0,"URL":"https:\/\/doi.org\/10.15223\/policy-017"},{"start":{"date-parts":[[2026,10,1]],"date-time":"2026-10-01T00:00:00Z","timestamp":1790812800000},"content-version":"stm-asf","delay-in-days":0,"URL":"https:\/\/doi.org\/10.15223\/policy-037"},{"start":{"date-parts":[[2026,10,1]],"date-time":"2026-10-01T00:00:00Z","timestamp":1790812800000},"content-version":"stm-asf","delay-in-days":0,"URL":"https:\/\/doi.org\/10.15223\/policy-012"},{"start":{"date-parts":[[2026,10,1]],"date-time":"2026-10-01T00:00:00Z","timestamp":1790812800000},"content-version":"stm-asf","delay-in-days":0,"URL":"https:\/\/doi.org\/10.15223\/policy-029"},{"start":{"date-parts":[[2026,10,1]],"date-time":"2026-10-01T00:00:00Z","timestamp":1790812800000},"content-version":"stm-asf","delay-in-days":0,"URL":"https:\/\/doi.org\/10.15223\/policy-004"}],"funder":[{"DOI":"10.13039\/501100000289","name":"Cancer Research UK","doi-asserted-by":"publisher","award":["SEBSTF-2025\/100007"],"award-info":[{"award-number":["SEBSTF-2025\/100007"]}],"id":[{"id":"10.13039\/501100000289","id-type":"DOI","asserted-by":"publisher"}]},{"DOI":"10.13039\/100023015","name":"Advanced Research Projects Agency for Health","doi-asserted-by":"publisher","award":["ADAPT OT 140D042590013"],"award-info":[{"award-number":["ADAPT OT 140D042590013"]}],"id":[{"id":"10.13039\/100023015","id-type":"DOI","asserted-by":"publisher"}]}],"content-domain":{"domain":["cell.com","elsevier.com","sciencedirect.com"],"crossmark-restriction":true},"short-container-title":["Molecular Cell"],"published-print":{"date-parts":[[2026,10]]},"DOI":"10.1016\/j.molcel.2026.09.030","type":"journal-article","created":{"date-parts":[[2026,10,7]],"date-time":"2026-10-07T14:54:13Z","timestamp":1791384853000},"page":"4097-4107","update-policy":"https:\/\/doi.org\/10.1016\/elsevier_cm_policy","source":"Crossref","is-referenced-by-count":0,"title":["Toward interpretable foundation models for molecular biology"],"prefix":"10.1016","volume":"86","author":[{"ORCID":"https:\/\/orcid.org\/0000-0002-9254-1247","authenticated-orcid":false,"given":"Mukund","family":"Varma","sequence":"first","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Trey","family":"Ideker","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]}],"member":"78","reference":[{"key":"10.1016\/j.molcel.2026.09.030_bib1","doi-asserted-by":"crossref","first-page":"860","DOI":"10.1038\/35057062","article-title":"Initial sequencing and analysis of the human genome","volume":"409","author":"Lander","year":"2001","journal-title":"Nature"},{"key":"10.1016\/j.molcel.2026.09.030_bib2","doi-asserted-by":"crossref","first-page":"2539","DOI":"10.1038\/s41592-025-02899-6","article-title":"Novae: a graph-based foundation model for spatial transcriptomics data","volume":"22","author":"Blampey","year":"2025","journal-title":"Nat. Methods"},{"key":"10.1016\/j.molcel.2026.09.030_bib3","series-title":"Proceedings of the IEEE\/CVF Conference on Computer Vision and Pattern Recognition (CVPR 2024)","first-page":"9632","article-title":"Transcriptomics-guided slide representation learning in computational pathology","author":"Jaume","year":"2024"},{"key":"10.1016\/j.molcel.2026.09.030_bib4","doi-asserted-by":"crossref","first-page":"D672","DOI":"10.1093\/nar\/gkad1025","article-title":"The Reactome Pathway Knowledgebase 2024","volume":"52","author":"Milacic","year":"2024","journal-title":"Nucleic Acids Res."},{"key":"10.1016\/j.molcel.2026.09.030_bib5","doi-asserted-by":"crossref","first-page":"417","DOI":"10.1016\/j.cels.2015.12.004","article-title":"The Molecular Signatures Database (MSigDB) hallmark gene set collection","volume":"1","author":"Liberzon","year":"2015","journal-title":"Cell Syst."},{"key":"10.1016\/j.molcel.2026.09.030_bib6","series-title":"Proceedings of the IEEE\/CVF Conference on Computer Vision and Pattern Recognition (CVPR 2024)","first-page":"11579","article-title":"Modeling dense multimodal interactions between biological pathways and histology for survival prediction","author":"Jaume","year":"2024"},{"key":"10.1016\/j.molcel.2026.09.030_bib7","doi-asserted-by":"crossref","first-page":"1084","DOI":"10.1038\/s41587-023-01940-3","article-title":"Supervised discovery of interpretable gene programs from single-cell data","volume":"42","author":"Kunes","year":"2024","journal-title":"Nat. Biotechnol."},{"key":"10.1016\/j.molcel.2026.09.030_bib8","doi-asserted-by":"crossref","DOI":"10.1158\/2159-8290.CD-25-1735","article-title":"A foundation model of cancer genotype enables precise predictions of therapeutic response","author":"Kong","year":"2026","journal-title":"Cancer Discov."},{"key":"10.1016\/j.molcel.2026.09.030_bib9","doi-asserted-by":"crossref","first-page":"583","DOI":"10.1038\/s41586-021-03819-2","article-title":"Highly accurate protein structure prediction with AlphaFold","volume":"596","author":"Jumper","year":"2021","journal-title":"Nature"},{"key":"10.1016\/j.molcel.2026.09.030_bib10","article-title":"Illuminating cell states by a comprehensive and interpretable single cell foundation model","volume":"17","author":"Wang","year":"2026","journal-title":"Nat. Commun."},{"key":"10.1016\/j.molcel.2026.09.030_bib11","doi-asserted-by":"crossref","first-page":"935","DOI":"10.1038\/s41592-026-03064-3","article-title":"Orthrus: toward evolutionary and functional RNA foundation models","volume":"23","author":"Fradkin","year":"2026","journal-title":"Nat. Methods"},{"key":"10.1016\/j.molcel.2026.09.030_bib12","doi-asserted-by":"crossref","first-page":"1470","DOI":"10.1038\/s41592-024-02201-0","article-title":"scGPT: toward building a foundation model for single-cell multi-omics using generative AI","volume":"21","author":"Cui","year":"2024","journal-title":"Nat. Methods"},{"key":"10.1016\/j.molcel.2026.09.030_bib13","article-title":"ChromBERT: a foundation model for learning interpretable representations for context-specific transcriptional regulatory networks","volume":"6","author":"Yu","year":"2026","journal-title":"Cell Genom."},{"key":"10.1016\/j.molcel.2026.09.030_bib14","article-title":"Multimodal learning enables chat-based exploration of single-cell data","author":"Schaefer","year":"2025","journal-title":"Nat. Biotechnol."},{"key":"10.1016\/j.molcel.2026.09.030_bib15","unstructured":"Ismail, A.A., Oikarinen, T., Wang, A., Adebayo, J., Stanton, S.D., Corrada Bravo, H., Cho, K., and Frey, N.C. (2025). Concept bottleneck language models for protein design. https:\/\/doi.org\/10.48550\/arXiv.2411.06090"},{"key":"10.1016\/j.molcel.2026.09.030_bib16","doi-asserted-by":"crossref","first-page":"17598","DOI":"10.1021\/jacs.4c07099","article-title":"MACE-OFF: short-range transferable machine learning force fields for organic molecules","volume":"147","author":"Kov\u00e1cs","year":"2025","journal-title":"J. Am. Chem. Soc."},{"key":"10.1016\/j.molcel.2026.09.030_bib17","doi-asserted-by":"crossref","first-page":"1256","DOI":"10.1038\/s42256-022-00580-7","article-title":"Large-scale chemical language representations capture molecular structure and properties","volume":"4","author":"Ross","year":"2022","journal-title":"Nat. Mach. Intell."},{"key":"10.1016\/j.molcel.2026.09.030_bib18","doi-asserted-by":"crossref","unstructured":"Su, J., Han, C., Zhou, Y., Shan, J., Zhou, X., and Yuan, F. (2024). SaProt: protein language modeling with structure-aware vocabulary. https:\/\/openreview.net\/forum?id=6MRm3G4NiU.","DOI":"10.1101\/2023.10.01.560349"},{"key":"10.1016\/j.molcel.2026.09.030_bib19","doi-asserted-by":"crossref","first-page":"616","DOI":"10.1038\/s41586-023-06139-9","article-title":"Transfer learning enables predictions in network biology","volume":"618","author":"Theodoris","year":"2023","journal-title":"Nature"},{"key":"10.1016\/j.molcel.2026.09.030_bib20","doi-asserted-by":"crossref","first-page":"1053","DOI":"10.1038\/s41592-018-0229-2","article-title":"Deep generative modeling for single-cell transcriptomics","volume":"15","author":"Lopez","year":"2018","journal-title":"Nat. Methods"},{"key":"10.1016\/j.molcel.2026.09.030_bib21","article-title":"The Gene Ontology knowledgebase in 2023","volume":"224","author":"Aleksander","year":"2023","journal-title":"Genetics"},{"key":"10.1016\/j.molcel.2026.09.030_bib22","doi-asserted-by":"crossref","first-page":"D638","DOI":"10.1093\/nar\/gkac1000","article-title":"The STRING database in 2023: protein-protein association networks and functional enrichment analyses for any sequenced genome of interest","volume":"51","author":"Szklarczyk","year":"2023","journal-title":"Nucleic Acids Res."},{"key":"10.1016\/j.molcel.2026.09.030_bib23","doi-asserted-by":"crossref","first-page":"290","DOI":"10.1038\/nmeth.4627","article-title":"Using deep learning to model the hierarchical structure and function of a cell","volume":"15","author":"Ma","year":"2018","journal-title":"Nat. Methods"},{"key":"10.1016\/j.molcel.2026.09.030_bib24","doi-asserted-by":"crossref","first-page":"348","DOI":"10.1038\/s41586-021-03922-4","article-title":"Biologically informed deep neural network for prostate cancer discovery","volume":"598","author":"Elmarakeby","year":"2021","journal-title":"Nature"},{"key":"10.1016\/j.molcel.2026.09.030_bib25","doi-asserted-by":"crossref","DOI":"10.1186\/s13059-020-02100-5","article-title":"Knowledge-primed neural networks enable biologically interpretable deep learning on single-cell sequencing data","volume":"21","author":"Fortelny","year":"2020","journal-title":"Genome Biol."},{"key":"10.1016\/j.molcel.2026.09.030_bib26","doi-asserted-by":"crossref","DOI":"10.1038\/s42003-021-02622-z","article-title":"GenNet framework: interpretable deep learning for predicting phenotypes from genetic data","volume":"4","author":"van Hilten","year":"2021","journal-title":"Commun. Biol."},{"key":"10.1016\/j.molcel.2026.09.030_bib27","doi-asserted-by":"crossref","first-page":"996","DOI":"10.1038\/s43018-024-00740-1","article-title":"A deep learning model of tumor cell architecture elucidates response and resistance to CDK4\/6 inhibitors","volume":"5","author":"Park","year":"2024","journal-title":"Nat. Cancer"},{"key":"10.1016\/j.molcel.2026.09.030_bib28","doi-asserted-by":"crossref","first-page":"493","DOI":"10.1038\/s41586-024-07487-w","article-title":"Accurate structure prediction of biomolecular interactions with AlphaFold 3","volume":"630","author":"Abramson","year":"2024","journal-title":"Nature"},{"key":"10.1016\/j.molcel.2026.09.030_bib29","series-title":"Proceedings of the Forty-first International Conference on Machine Learning (ICML 2024)","first-page":"43632","article-title":"Caduceus: bi-directional equivariant long-range DNA sequence modeling","volume":"235","author":"Schiff","year":"2024"},{"key":"10.1016\/j.molcel.2026.09.030_bib30","doi-asserted-by":"crossref","DOI":"10.1186\/s13059-025-03574-x","article-title":"Zero-shot evaluation reveals limitations of single-cell foundation models","volume":"26","author":"Kedzierska","year":"2025","journal-title":"Genome Biol."},{"key":"10.1016\/j.molcel.2026.09.030_bib31","doi-asserted-by":"crossref","DOI":"10.1038\/s41587-026-03166-5","article-title":"Experiment-guided AlphaFold3 resolves measurement-consistent protein ensembles","author":"Maddipatla","year":"2026","journal-title":"Nat. Biotechnol."},{"key":"10.1016\/j.molcel.2026.09.030_bib32","doi-asserted-by":"crossref","first-page":"3418","DOI":"10.1093\/bioinformatics\/btaa169","article-title":"Interpretable factor models of single-cell RNA-seq via variational autoencoders","volume":"36","author":"Svensson","year":"2020","journal-title":"Bioinformatics"},{"key":"10.1016\/j.molcel.2026.09.030_bib33","doi-asserted-by":"crossref","first-page":"128","DOI":"10.1016\/j.cels.2020.11.013","article-title":"CellBox: interpretable machine learning for perturbation biology with application to the design of cancer combination therapy","volume":"12","author":"Yuan","year":"2021","journal-title":"Cell Syst."},{"key":"10.1016\/j.molcel.2026.09.030_bib34","doi-asserted-by":"crossref","first-page":"125","DOI":"10.1038\/s41576-022-00532-2","article-title":"Obtaining genetics insights from deep learning via explainable artificial intelligence","volume":"24","author":"Novakovsky","year":"2023","journal-title":"Nat. Rev. Genet."},{"key":"10.1016\/j.molcel.2026.09.030_bib35","doi-asserted-by":"crossref","first-page":"1454","DOI":"10.1038\/s41592-024-02359-7","article-title":"Applying interpretable machine learning in computational biology\u2014pitfalls, recommendations and opportunities for new developments","volume":"21","author":"Chen","year":"2024","journal-title":"Nat. Methods"},{"key":"10.1016\/j.molcel.2026.09.030_bib36","doi-asserted-by":"crossref","DOI":"10.1073\/pnas.2506316122","article-title":"Sparse autoencoders uncover biologically interpretable features in protein language model representations","volume":"122","author":"Gujral","year":"2025","journal-title":"Proc. Natl. Acad. Sci. USA"},{"key":"10.1016\/j.molcel.2026.09.030_bib37","doi-asserted-by":"crossref","DOI":"10.1186\/s13321-025-01079-0","article-title":"Measuring chemical LLM robustness to molecular representations: a SMILES variation-based framework","volume":"17","author":"Ganeeva","year":"2025","journal-title":"J. Cheminform."},{"key":"10.1016\/j.molcel.2026.09.030_bib38","series-title":"Proceedings of the Fortieth International Conference on Machine Learning (ICML 2023)","first-page":"10929","article-title":"RankMe: assessing the downstream performance of pretrained self-supervised representations by their rank","volume":"202","author":"Garrido","year":"2023"},{"key":"10.1016\/j.molcel.2026.09.030_bib39","doi-asserted-by":"crossref","first-page":"1657","DOI":"10.1038\/s41592-025-02772-6","article-title":"Deep-learning-based gene perturbation effect prediction does not yet outperform simple linear baselines","volume":"22","author":"Ahlmann-Eltze","year":"2025","journal-title":"Nat. Methods"},{"key":"10.1016\/j.molcel.2026.09.030_bib40","doi-asserted-by":"crossref","DOI":"10.1136\/bmj-2023-078378","article-title":"TRIPOD+AI statement: updated guidance for reporting clinical prediction models that use regression or machine learning methods","volume":"385","author":"Collins","year":"2024","journal-title":"BMJ"}],"container-title":["Molecular Cell"],"original-title":[],"language":"en","link":[{"URL":"https:\/\/api.elsevier.com\/content\/article\/PII:S1097276526006738?httpAccept=text\/xml","content-type":"text\/xml","content-version":"vor","intended-application":"text-mining"},{"URL":"https:\/\/api.elsevier.com\/content\/article\/PII:S1097276526006738?httpAccept=text\/plain","content-type":"text\/plain","content-version":"vor","intended-application":"text-mining"}],"deposited":{"date-parts":[[2026,10,15]],"date-time":"2026-10-15T15:21:46Z","timestamp":1792077706000},"score":1,"resource":{"primary":{"URL":"https:\/\/linkinghub.elsevier.com\/retrieve\/pii\/S1097276526006738"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2026,10]]},"references-count":40,"journal-issue":{"issue":"20","published-print":{"date-parts":[[2026,10]]}},"alternative-id":["S1097276526006738"],"URL":"https:\/\/doi.org\/10.1016\/j.molcel.2026.09.030","relation":{},"ISSN":["1097-2765"],"issn-type":[{"value":"1097-2765","type":"print"}],"subject":[],"published":{"date-parts":[[2026,10]]},"assertion":[{"value":"Elsevier","name":"publisher","label":"This article is maintained by"},{"value":"Toward interpretable foundation models for molecular biology","name":"articletitle","label":"Article Title"},{"value":"Molecular Cell","name":"journaltitle","label":"Journal Title"},{"value":"https:\/\/doi.org\/10.1016\/j.molcel.2026.09.030","name":"articlelink","label":"CrossRef DOI link to publisher maintained version"},{"value":"article","name":"content_type","label":"Content Type"},{"value":"\u00a9 2026 Elsevier Inc. All rights are reserved, including those for text and data mining, AI training, and similar technologies.","name":"copyright","label":"Copyright"}]}}