{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2026,10,26]],"date-time":"2026-10-26T18:17:13Z","timestamp":1793038633470,"version":"4.5.0"},"reference-count":43,"publisher":"Oxford University Press (OUP)","issue":"5","license":[{"start":{"date-parts":[[2025,4,15]],"date-time":"2025-04-15T00:00:00Z","timestamp":1744675200000},"content-version":"vor","delay-in-days":0,"URL":"https:\/\/creativecommons.org\/licenses\/by\/4.0\/"}],"content-domain":{"domain":[],"crossmark-restriction":false},"short-container-title":[],"published-print":{"date-parts":[[2025,5,6]]},"abstract":"<jats:title>Abstract<\/jats:title>\n                  <jats:sec>\n                    <jats:title>Motivation<\/jats:title>\n                    <jats:p>Understanding the protein sequence\u2013function relationship is essential for advancing protein biology and engineering. However, &amp;lt;1% of known protein sequences have human-verified functions. While deep-learning methods have demonstrated promise for protein\u2013function prediction, current models are limited to predicting only those functions on which they were trained.<\/jats:p>\n                  <\/jats:sec>\n                  <jats:sec>\n                    <jats:title>Results<\/jats:title>\n                    <jats:p>Here, we introduce ProtNote, a multimodal deep-learning model that leverages free-form text to enable both supervised and zero-shot protein\u2013function prediction. ProtNote not only maintains near state-of-the-art performance for annotations in its training set but also generalizes to unseen and novel functions in zero-shot test settings. ProtNote demonstrates superior performance in the prediction of novel Gene Ontology annotations and Enzyme Commission numbers compared to baseline models by capturing nuanced sequence\u2013function relationships that unlock a range of biological use cases inaccessible to prior models. We envision that ProtNote will enhance protein\u2013function discovery by enabling scientists to use free text inputs without restriction to predefined labels\u2014a necessary capability for navigating the dynamic landscape of protein biology.<\/jats:p>\n                  <\/jats:sec>\n                  <jats:sec>\n                    <jats:title>Availability and Implementation<\/jats:title>\n                    <jats:p>The code is available on GitHub: https:\/\/github.com\/microsoft\/protnote; model weights, datasets, and evaluation metrics are provided via Zenodo: https:\/\/zenodo.org\/records\/13897920.<\/jats:p>\n                  <\/jats:sec>","DOI":"10.1093\/bioinformatics\/btaf170","type":"journal-article","created":{"date-parts":[[2025,4,12]],"date-time":"2025-04-12T16:59:16Z","timestamp":1744477156000},"source":"Crossref","is-referenced-by-count":20,"title":["ProtNote: a multimodal method for protein\u2013function annotation"],"prefix":"10.1093","volume":"41","author":[{"given":"Samir","family":"Char","sequence":"first","affiliation":[{"name":"Microsoft Cloud & AI, Microsoft , Redmond, WA 98052,","place":["United States"]}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Nathaniel","family":"Corley","sequence":"additional","affiliation":[{"name":"Microsoft Cloud & AI, Health & Life Sciences, Microsoft , Redmond, WA 98052,","place":["United States"]}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Sarah","family":"Alamdari","sequence":"additional","affiliation":[{"name":"Microsoft Research, Microsoft , 1 Memorial Dr , Cambridge, MA 02142,","place":["United States"]}],"role":[{"vocabulary":"crossref","role":"author"}]},{"ORCID":"https:\/\/orcid.org\/0000-0001-9045-6826","authenticated-orcid":false,"given":"Kevin K","family":"Yang","sequence":"additional","affiliation":[{"name":"Microsoft Research, Microsoft , 1 Memorial Dr , Cambridge, MA 02142,","place":["United States"]}],"role":[{"vocabulary":"crossref","role":"author"}]},{"ORCID":"https:\/\/orcid.org\/0000-0002-8601-6040","authenticated-orcid":false,"given":"Ava P","family":"Amini","sequence":"additional","affiliation":[{"name":"Microsoft Research, Microsoft , 1 Memorial Dr , Cambridge, MA 02142,","place":["United States"]}],"role":[{"vocabulary":"crossref","role":"author"}]}],"member":"286","published-online":{"date-parts":[[2025,4,15]]},"reference":[{"key":"2025051307513304600_btaf170-B1","doi-asserted-by":"publisher","DOI":"10.48550\/arXiv.1803.08375","article-title":"Deep learning using rectified linear units (ReLU)","author":"Agarap","year":"2018","journal-title":"CoRR Abs\/1803.08375"},{"key":"2025051307513304600_btaf170-B2","doi-asserted-by":"crossref","first-page":"403","DOI":"10.1016\/S0022-2836(05)80360-2","article-title":"Basic local alignment search tool","volume":"215","author":"Altschul","year":"1990","journal-title":"J Mol Biol"},{"key":"2025051307513304600_btaf170-B3","doi-asserted-by":"publisher","first-page":"25","DOI":"10.1038\/75556","article-title":"Gene ontology: tool for the unification of biology","volume":"25","author":"Ashburner","year":"2000","journal-title":"Nat Genet"},{"key":"2025051307513304600_btaf170-B4","doi-asserted-by":"publisher","article-title":"Annotating the Microbial Dark Matter with HiFi-NN","journal-title":"ISCIENCE","DOI":"10.1016\/j.isci.2025.112480"},{"key":"2025051307513304600_btaf170-B5","doi-asserted-by":"publisher","first-page":"304","DOI":"10.1093\/nar\/28.1.304","article-title":"The ENZYME database in 2000","volume":"28","author":"Bairoch","year":"2000","journal-title":"Nucleic Acids Res"},{"key":"2025051307513304600_btaf170-B6","doi-asserted-by":"publisher","first-page":"932","DOI":"10.1038\/s41587-021-01179-w","article-title":"Using deep learning to annotate the protein universe","volume":"40","author":"Bileschi","year":"2022","journal-title":"Nat Biotechnol"},{"key":"2025051307513304600_btaf170-B7","first-page":"89","article-title":"UniProtKB\/Swiss-Prot","volume":"406","author":"Boutet","year":"2007","journal-title":"Methods Mol Biol"},{"key":"2025051307513304600_btaf170-B8","doi-asserted-by":"publisher","first-page":"2102","DOI":"10.1093\/BIOINFORMATICS\/BTAC020","article-title":"ProteinBERT: a universal deep-learning model of protein sequence and function","volume":"38","author":"Brandes","year":"2022","journal-title":"Bioinformatics"},{"key":"2025051307513304600_btaf170-B9","doi-asserted-by":"publisher","first-page":"iyad031","DOI":"10.1093\/genetics\/iyad031","article-title":"The gene ontology knowledgebase in 2023","volume":"224","author":"Gene Ontology Consortium","year":"2023","journal-title":"Genetics"},{"key":"2025051307513304600_btaf170-B10","doi-asserted-by":"crossref","first-page":"1","DOI":"10.1007\/978-1-61779-921-1_1","article-title":"Therapeutic proteins","volume":"899","author":"Dimitrov","year":"2012","journal-title":"Methods Mol Biol"},{"key":"2025051307513304600_btaf170-B11","doi-asserted-by":"publisher","first-page":"e1002195","DOI":"10.1371\/journal.pcbi.1002195","article-title":"Accelerated profile HMM searches","volume":"7","author":"Eddy","year":"2011","journal-title":"PLoS Comput Biol"},{"key":"2025051307513304600_btaf170-B12","author":"Gane","year":"2025"},{"key":"2025051307513304600_btaf170-B13","doi-asserted-by":"publisher","first-page":"3784","DOI":"10.1093\/nar\/gkg563","article-title":"ExPASy: the proteomics server for in-depth protein knowledge and analysis","volume":"31","author":"Gasteiger","year":"2003","journal-title":"Nucleic Acids Res"},{"key":"2025051307513304600_btaf170-B14","doi-asserted-by":"publisher","first-page":"3168","DOI":"10.1038\/s41467-021-23303-9","article-title":"Structure-based protein function prediction using graph convolutional networks","volume":"12.1","author":"Gligorijevi\u0107","year":"2021","journal-title":"Nat Commun"},{"key":"2025051307513304600_btaf170-B15","doi-asserted-by":"publisher","first-page":"1","DOI":"10.1145\/3458754","article-title":"Domain-Specific language model pretraining for biomedical natural language processing","volume":"3","author":"Gu","year":"2022","journal-title":"ACM Trans Comput Healthcare"},{"key":"2025051307513304600_btaf170-B27","doi-asserted-by":"publisher","first-page":"57","DOI":"10.1002\/(SICI)1098-1004(200001)15:1&lt;57::AID-HUMU12&gt;3.0.CO;2-G","article-title":"Online mendelian inheritance in man (OMIM)","volume":"15","author":"Hamosh","year":"2000","journal-title":"Hum Mutat"},{"key":"2025051307513304600_btaf170-B16","doi-asserted-by":"crossref","first-page":"10915","DOI":"10.1073\/pnas.89.22.10915","article-title":"Amino acid substitution matrices from protein blocks","volume":"89","author":"Henikoff","year":"1992","journal-title":"Proc Natl Acad Sci USA"},{"key":"2025051307513304600_btaf170-B17","first-page":"448","author":"Ioffe"},{"key":"2025051307513304600_btaf170-B18","doi-asserted-by":"publisher","author":"Kingma","DOI":"10.48550\/arXiv.1412.6980"},{"key":"2025051307513304600_btaf170-B19","doi-asserted-by":"publisher","author":"Kipf","DOI":"10.48550\/arXiv.1609.02907"},{"key":"2025051307513304600_btaf170-B20","doi-asserted-by":"publisher","first-page":"1501","DOI":"10.1006\/jmbi1994.1104","article-title":"Hidden Markov models in computational biology. Applications to protein modeling","volume":"235","author":"Krogh","year":"1994","journal-title":"J Mol Biol"},{"key":"2025051307513304600_btaf170-B21","doi-asserted-by":"publisher","first-page":"1187","DOI":"10.1093\/BIOINFORMATICS\/BTAA763","article-title":"DeepGOPlus: improved protein function prediction from sequence","volume":"37","author":"Kulmanov","year":"2021","journal-title":"Bioinformatics"},{"key":"2025051307513304600_btaf170-B22","doi-asserted-by":"publisher","first-page":"i238","DOI":"10.1093\/bioinformatics\/btac256","article-title":"DeepGOZero: improving protein function prediction from sequence and zero-shot learning based on ontology axioms","volume":"38","author":"Kulmanov","year":"2022","journal-title":"Bioinformatics"},{"key":"2025051307513304600_btaf170-B23","doi-asserted-by":"publisher","first-page":"21","DOI":"10.1038\/nrd2399","article-title":"Protein therapeutics: a summary and pharmacological classification","volume":"7","author":"Leader","year":"2008","journal-title":"Nat Rev Drug Discov"},{"key":"2025051307513304600_btaf170-B24","doi-asserted-by":"publisher","author":"Lin","DOI":"10.1109\/ICCV.2017.324"},{"key":"2025051307513304600_btaf170-B25","doi-asserted-by":"publisher","DOI":"10.1093\/BIB\/BBAC409","article-title":"BioGPT: generative pre-trained transformer for biomedical text generation and mining","volume":"23","author":"Luo","year":"2022","journal-title":"Briefings Bioinform"},{"key":"2025051307513304600_btaf170-B26","doi-asserted-by":"publisher","first-page":"77","DOI":"10.3390\/synbio1010006","article-title":"Proteins in synthetic biology with agricultural and environmental applications","volume":"1","author":"Mann","year":"2023","journal-title":"SynBio"},{"key":"2025051307513304600_btaf170-B28","author":"Micikevicius"},{"key":"2025051307513304600_btaf170-B29","doi-asserted-by":"publisher","first-page":"D412","DOI":"10.1093\/nar\/gkaa913","article-title":"Pfam: the protein families database in 2021","volume":"49","author":"Mistry","year":"2021","journal-title":"Nucleic Acids Res"},{"key":"2025051307513304600_btaf170-B30","doi-asserted-by":"publisher","first-page":"221","DOI":"10.1038\/nmeth.2340","article-title":"A large-scale evaluation of computational protein function prediction","volume":"10","author":"Radivojac","year":"2013","journal-title":"Nat Methods"},{"key":"2025051307513304600_btaf170-B31","doi-asserted-by":"crossref","first-page":"16","DOI":"10.17113\/ftb.56.01.18.5491","article-title":"Applications of microbial enzymes in food industry","volume":"56","author":"Raveendran","year":"2018","journal-title":"Food Technol Biotechnol"},{"key":"2025051307513304600_btaf170-B32","doi-asserted-by":"publisher","first-page":"e80942","DOI":"10.7554\/eLife.80942","article-title":"ProteInfer, deep neural networks for protein functional inference","volume":"12","author":"Sanderson","year":"2023","journal-title":"Elife"},{"key":"2025051307513304600_btaf170-B33","doi-asserted-by":"publisher","first-page":"7407","DOI":"10.1038\/s41467-024-51844-2","article-title":"Fine-tuning protein language models boosts predictions across diverse tasks","volume":"15","author":"Schmirler","year":"2024","journal-title":"Nat Commun"},{"key":"2025051307513304600_btaf170-B34","doi-asserted-by":"crossref","first-page":"47","DOI":"10.1093\/nar\/30.1.47","article-title":"BRENDA, enzyme data and metabolic information","volume":"30","author":"Schomburg","year":"2002","journal-title":"Nucleic Acids Res"},{"key":"2025051307513304600_btaf170-B35","doi-asserted-by":"publisher","year":"2024","DOI":"10.1101\/2024.05.30.596539"},{"key":"2025051307513304600_btaf170-B36","doi-asserted-by":"publisher","DOI":"10.1093\/nar\/gks1067","article-title":"New and continuing developments at PROSITE","volume":"41","author":"Sigrist","year":"2012","journal-title":"Nucl Acids Res"},{"key":"2025051307513304600_btaf170-B37","doi-asserted-by":"publisher","first-page":"D523","DOI":"10.1093\/nar\/gkac1052","article-title":"UniProt: the universal protein knowledgebase in 2023","volume":"51","author":"The UniProt Consortium","year":"2022","journal-title":"Nucleic Acids Research"},{"key":"2025051307513304600_btaf170-B38","doi-asserted-by":"publisher","DOI":"10.48550\/ARXIV.2402.05672","article-title":"Multilingual E5 text embeddings: a technical report","author":"Wang","year":"2024"},{"key":"2025051307513304600_btaf170-B39","doi-asserted-by":"publisher","author":"Wu","year":"2024","DOI":"10.1101\/2024.05.14.594226"},{"key":"2025051307513304600_btaf170-B40","doi-asserted-by":"publisher","first-page":"279","DOI":"10.1007\/978-3-031-04749-7_17","volume-title":"Research in computational molecular biology","author":"Xu","year":"2022"},{"key":"2025051307513304600_btaf170-B41","doi-asserted-by":"publisher","first-page":"1358","DOI":"10.1126\/science.adf2465","article-title":"Enzyme function prediction using contrastive learning","volume":"379","author":"Yu","year":"2023","journal-title":"Science"},{"key":"2025051307513304600_btaf170-B42","doi-asserted-by":"publisher","author":"Zhang","DOI":"10.1109\/CVPR52688.2022.01616"},{"key":"2025051307513304600_btaf170-B43","doi-asserted-by":"publisher","first-page":"e2406285121","DOI":"10.1073\/pnas.2406285121","article-title":"Protein language models learn evolutionary statistics of interacting sequence motifs","volume":"121","author":"Zhang","year":"2024","journal-title":"Proc Natl Acad Sci U S A"}],"container-title":["Bioinformatics"],"original-title":[],"language":"en","link":[{"URL":"https:\/\/academic.oup.com\/bioinformatics\/advance-article-pdf\/doi\/10.1093\/bioinformatics\/btaf170\/62932092\/btaf170.pdf","content-type":"application\/pdf","content-version":"am","intended-application":"syndication"},{"URL":"https:\/\/academic.oup.com\/bioinformatics\/article-pdf\/41\/5\/btaf170\/62932092\/btaf170.pdf","content-type":"application\/pdf","content-version":"vor","intended-application":"syndication"},{"URL":"https:\/\/academic.oup.com\/bioinformatics\/article-pdf\/41\/5\/btaf170\/62932092\/btaf170.pdf","content-type":"unspecified","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2025,5,13]],"date-time":"2025-05-13T07:51:41Z","timestamp":1747122701000},"score":1,"resource":{"primary":{"URL":"https:\/\/academic.oup.com\/bioinformatics\/article\/doi\/10.1093\/bioinformatics\/btaf170\/8113843"}},"subtitle":[],"editor":[{"given":"Pier Luigi","family":"Martelli","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"editor"}]}],"short-title":[],"issued":{"date-parts":[[2025,4,15]]},"references-count":43,"journal-issue":{"issue":"5","published-print":{"date-parts":[[2025,5,6]]}},"URL":"https:\/\/doi.org\/10.1093\/bioinformatics\/btaf170","relation":{"has-preprint":[{"id-type":"doi","id":"10.1101\/2024.10.17.618952","asserted-by":"object"}]},"ISSN":["1367-4811"],"issn-type":[{"value":"1367-4811","type":"electronic"}],"subject":[],"published-other":{"date-parts":[[2025,5]]},"published":{"date-parts":[[2025,4,15]]},"article-number":"btaf170"}}