{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"institution":[{"name":"bioRxiv"}],"indexed":{"date-parts":[[2026,4,11]],"date-time":"2026-04-11T08:05:14Z","timestamp":1775894714368,"version":"3.50.1"},"posted":{"date-parts":[[2025,3,13]]},"group-title":"Bioinformatics","reference-count":39,"publisher":"openRxiv","license":[{"start":{"date-parts":[[2025,3,13]],"date-time":"2025-03-13T00:00:00Z","timestamp":1741824000000},"content-version":"vor","delay-in-days":0,"URL":"http:\/\/creativecommons.org\/licenses\/by-nc\/4.0\/"}],"content-domain":{"domain":[],"crossmark-restriction":false},"short-container-title":[],"accepted":{"date-parts":[[2025,3,13]]},"abstract":"<jats:p>The early detection of high-fitness viral variants is critical for pandemic response, yet limited experimental resources at the onset of variant emergence hinder effective identification. To address this, we introduce an active learning framework that integrates protein language model ESM3, Gaussian process with uncertainty estimation, and a bio-physical model to predict the fitness of novel variants in a few-shot learning setting. By benchmarking on past SARS-CoV-2 data, we demonstrate that our methods accelerates the identification of high-fitness variants by up to fivefold compared to random sampling while requiring experimental characterization of fewer than 1% of possible variants. We also demonstrate that our framework benchmarked on deep mutational scans effectively identifies sites that are frequently mutated during natural viral evolution with a predictive advantage of up to two years compared to baseline strategies, particularly those enabling antibody escape while preserving ACE2 binding. Through systematic analysis of different acquisition strategies, we show that incorporating uncertainty in variant selection enables broader exploration of the sequence landscape, leading to the discovery of evolutionarily distant but potentially dangerous variants. Our results suggest that this framework could serve as an effective early warning system for identifying concerning SARS-CoV-2 variants and potentially emerging viruses with pandemic potential before they achieve widespread circulation.<\/jats:p>","DOI":"10.1101\/2025.03.12.642881","type":"posted-content","created":{"date-parts":[[2025,3,14]],"date-time":"2025-03-14T03:20:21Z","timestamp":1741922421000},"source":"Crossref","is-referenced-by-count":2,"title":["Few-Shot Viral Variant Detection via Bayesian Active Learning and Biophysics"],"prefix":"10.64898","author":[{"ORCID":"https:\/\/orcid.org\/0009-0002-2359-5185","authenticated-orcid":false,"given":"Marian","family":"Huot","sequence":"first","affiliation":[{"name":"Department of Chemistry and Chemical Biology, Harvard University, Cambridge, MA"},{"name":"Laboratory of Physics of the Ecole Normale Sup\u00e9rieure, CNRS UMR 8023 and PSL Research, Sorbonne Universit\u00e9"}],"role":[{"role":"author","vocabulary":"crossref"}]},{"ORCID":"https:\/\/orcid.org\/0000-0002-5503-1838","authenticated-orcid":false,"given":"Dianzhuo","family":"Wang","sequence":"additional","affiliation":[{"name":"Department of Chemistry and Chemical Biology, Harvard University, Cambridge, MA"},{"name":"John A. Paulson School of Engineering and Applied Sciences, Harvard University, Cambridge, MA"}],"role":[{"role":"author","vocabulary":"crossref"}]},{"ORCID":"https:\/\/orcid.org\/0000-0003-3308-2869","authenticated-orcid":false,"given":"Jiacheng","family":"Liu","sequence":"additional","affiliation":[{"name":"University of Washington, Seattle, WA"}],"role":[{"role":"author","vocabulary":"crossref"}]},{"ORCID":"https:\/\/orcid.org\/0000-0002-4769-2265","authenticated-orcid":false,"given":"Eugene","family":"Shakhnovich","sequence":"additional","affiliation":[{"name":"Department of Chemistry and Chemical Biology, Harvard University, Cambridge, MA"}],"role":[{"role":"author","vocabulary":"crossref"}]}],"member":"54368","reference":[{"key":"2026041100373508000_2025.03.12.642881v1.1","doi-asserted-by":"publisher","DOI":"10.1038\/s41586-023-06487-6"},{"key":"2026041100373508000_2025.03.12.642881v1.2","doi-asserted-by":"crossref","unstructured":"Allman, B. , Vieira, L. , Diaz, D. J. , and Wilke, C. O. A systematic evaluation of the languageof-viral-escape model using multiple machine learning frameworks, (2024). URL http:\/\/biorxiv.org\/lookup\/doi\/10.1101\/2024.09.04.611278.","DOI":"10.1101\/2024.09.04.611278"},{"key":"2026041100373508000_2025.03.12.642881v1.3","doi-asserted-by":"publisher","DOI":"10.7554\/eLife.70658"},{"key":"2026041100373508000_2025.03.12.642881v1.4","doi-asserted-by":"publisher","DOI":"10.1038\/s41579-022-00841-7"},{"key":"2026041100373508000_2025.03.12.642881v1.5","doi-asserted-by":"publisher","DOI":"10.1002\/pro.2915"},{"key":"2026041100373508000_2025.03.12.642881v1.6","unstructured":"Donmez, P. , Carbonell, J. G. , and Bennett, P. N. Dual strategy active learning. In Kok, J.N. , Koronacki, J. , Mantaras, R.L.d. , Matwin, S. , Mladeni\u010d, D. , and Skowron, A. , editors, Machine Learning: ECML 2007, pages 116\u2013127, Berlin, Heidelberg, (2007). Springer Berlin Heidelberg. ISBN 978-3-540-74958-5."},{"key":"2026041100373508000_2025.03.12.642881v1.7","doi-asserted-by":"publisher","DOI":"10.1038\/s41587-020-0521-4"},{"key":"2026041100373508000_2025.03.12.642881v1.8","doi-asserted-by":"publisher","DOI":"10.1002\/gch2.1018"},{"key":"2026041100373508000_2025.03.12.642881v1.9","doi-asserted-by":"publisher","DOI":"10.1002\/gch2.1018"},{"key":"2026041100373508000_2025.03.12.642881v1.10","unstructured":"Gessner, A. , Ober, S. W. , Vickery, O. , Ogli\u0107, D. , and U\u00e7ar, T. Active learning for affinity prediction of antibodies, (2024). URL https:\/\/arxiv.org\/abs\/2406.07263."},{"key":"2026041100373508000_2025.03.12.642881v1.11","doi-asserted-by":"publisher","DOI":"10.1039\/D0SC06805E"},{"key":"2026041100373508000_2025.03.12.642881v1.12","doi-asserted-by":"crossref","unstructured":"Hayes, T. , Rao, R. , Akin, H. , Sofroniew, N. J. , Oktay, D. , Lin, Z. , Verkuil, R. , Tran, V. Q. , Deaton, J. , Wiggert, M. , Badkundri, R. , Shafkat, I. , Gong, J. , Derry, A. , Molina, R. S. , Thomas, N. , Khan, Y. , Mishra, C. , Kim, C. , Bartie, L. J. , Nemeth, M. , Hsu, P. D. , Sercu, T. , Candido, S. , and Rives, A. Simulating 500 million years of evolution with a language model, (2024). URL http:\/\/biorxiv.org\/lookup\/doi\/10.1101\/2024.07.01.600583.","DOI":"10.1101\/2024.07.01.600583"},{"key":"2026041100373508000_2025.03.12.642881v1.13","doi-asserted-by":"publisher","DOI":"10.1016\/j.cels.2020.09.007"},{"key":"2026041100373508000_2025.03.12.642881v1.14","doi-asserted-by":"crossref","unstructured":"Ito, J. , Strange, A. , Liu, W. , Joas, G. , Lytras, S. , The Genotype to Phenotype Japan (G2P-Japan) Consortium, and Sato, K. A Protein Language Model for Exploring Viral Fitness Landscapes, (2024). URL http:\/\/biorxiv.org\/lookup\/doi\/10.1101\/2024.03.15.584819.","DOI":"10.1101\/2024.03.15.584819"},{"key":"2026041100373508000_2025.03.12.642881v1.15","doi-asserted-by":"publisher","DOI":"10.1093\/molbev\/mst010"},{"key":"2026041100373508000_2025.03.12.642881v1.16","doi-asserted-by":"publisher","DOI":"10.1021\/acs.jctc.2c00752"},{"key":"2026041100373508000_2025.03.12.642881v1.17","doi-asserted-by":"publisher","DOI":"10.1126\/science.ade2574"},{"key":"2026041100373508000_2025.03.12.642881v1.18","unstructured":"Loux, T. , Wang, D. , and Shakhnovich, E. I. Does structural information improve esm3 for protein binding affinity prediction?"},{"key":"2026041100373508000_2025.03.12.642881v1.19","doi-asserted-by":"publisher","DOI":"10.1126\/scitranslmed.abk3445"},{"key":"2026041100373508000_2025.03.12.642881v1.20","doi-asserted-by":"crossref","unstructured":"Meier, J. , Rao, R. , Verkuil, R. , Liu, J. , Sercu, T. , and Rives, A. Language models enable zero-shot prediction of the effects of mutations on protein function, (2021). URL http:\/\/biorxiv.org\/lookup\/doi\/10.1101\/2021.07.09.450648.","DOI":"10.1101\/2021.07.09.450648"},{"key":"2026041100373508000_2025.03.12.642881v1.21","doi-asserted-by":"publisher","DOI":"10.1016\/j.cell.2023.09.022"},{"key":"2026041100373508000_2025.03.12.642881v1.22","doi-asserted-by":"publisher","DOI":"10.1038\/s41467-022-34506-z"},{"key":"2026041100373508000_2025.03.12.642881v1.23","doi-asserted-by":"crossref","unstructured":"Moulana, A. , Dupic, T. , Phillips, A. M. , Chang, J. , Roffler, A. A. , Greaney, A. J. , Starr, T. N. , Bloom, J. D. , and Desai, M. M. (2023). The landscape of antibody binding affinity in SARS-CoV-2 omicron BA.1 evolution. Elife, 12.","DOI":"10.7554\/eLife.83442"},{"key":"2026041100373508000_2025.03.12.642881v1.24","doi-asserted-by":"publisher","DOI":"10.1126\/science.abl6251"},{"key":"2026041100373508000_2025.03.12.642881v1.25","doi-asserted-by":"publisher","DOI":"10.1126\/science.abm1208"},{"key":"2026041100373508000_2025.03.12.642881v1.26","doi-asserted-by":"publisher","DOI":"10.1038\/s41467-021-21118-2"},{"key":"2026041100373508000_2025.03.12.642881v1.27","doi-asserted-by":"publisher","DOI":"10.1073\/pnas.1215251110"},{"key":"2026041100373508000_2025.03.12.642881v1.28","doi-asserted-by":"publisher","DOI":"10.1093\/molbev\/msy131"},{"key":"2026041100373508000_2025.03.12.642881v1.29","doi-asserted-by":"publisher","DOI":"10.1142\/S0129065704001899"},{"key":"2026041100373508000_2025.03.12.642881v1.30","doi-asserted-by":"publisher","DOI":"10.1016\/j.cell.2020.08.012"},{"key":"2026041100373508000_2025.03.12.642881v1.31","doi-asserted-by":"publisher","DOI":"10.1038\/s41586-021-03807-6"},{"key":"2026041100373508000_2025.03.12.642881v1.32","doi-asserted-by":"publisher","DOI":"10.1126\/science.abf9302"},{"key":"2026041100373508000_2025.03.12.642881v1.33","doi-asserted-by":"publisher","DOI":"10.1016\/j.cell.2022.06.005"},{"key":"2026041100373508000_2025.03.12.642881v1.34","doi-asserted-by":"publisher","DOI":"10.1016\/j.bpj.2023.10.026"},{"key":"2026041100373508000_2025.03.12.642881v1.35","doi-asserted-by":"publisher","DOI":"10.1073\/PNAS.2314518121"},{"key":"2026041100373508000_2025.03.12.642881v1.36","doi-asserted-by":"publisher","DOI":"10.1016\/j.cell.2022.12.018"},{"key":"2026041100373508000_2025.03.12.642881v1.37","doi-asserted-by":"publisher","DOI":"10.1016\/j.molcel.2024.06.028"},{"key":"2026041100373508000_2025.03.12.642881v1.38","doi-asserted-by":"publisher","DOI":"10.2139\/ssrn.3639618"},{"key":"2026041100373508000_2025.03.12.642881v1.39","doi-asserted-by":"publisher","DOI":"10.1038\/nature13087"}],"container-title":[],"original-title":[],"link":[{"URL":"https:\/\/syndication.highwire.org\/content\/doi\/10.1101\/2025.03.12.642881","content-type":"unspecified","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2026,4,11]],"date-time":"2026-04-11T07:37:45Z","timestamp":1775893065000},"score":1,"resource":{"primary":{"URL":"http:\/\/biorxiv.org\/lookup\/doi\/10.1101\/2025.03.12.642881"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2025,3,13]]},"references-count":39,"URL":"https:\/\/doi.org\/10.1101\/2025.03.12.642881","relation":{},"subject":[],"published":{"date-parts":[[2025,3,13]]},"subtype":"preprint"}}