{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2026,3,10]],"date-time":"2026-03-10T22:36:31Z","timestamp":1773182191555,"version":"3.50.1"},"reference-count":54,"publisher":"Oxford University Press (OUP)","issue":"5","license":[{"start":{"date-parts":[[2025,5,27]],"date-time":"2025-05-27T00:00:00Z","timestamp":1748304000000},"content-version":"vor","delay-in-days":26,"URL":"https:\/\/creativecommons.org\/licenses\/by\/4.0\/"}],"funder":[{"DOI":"10.13039\/501100003977","name":"Israel Science foundation","doi-asserted-by":"publisher","award":["3486\/20"],"award-info":[{"award-number":["3486\/20"]}],"id":[{"id":"10.13039\/501100003977","id-type":"DOI","asserted-by":"publisher"}]}],"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>Protein\u2013protein interactions (PPIs) govern virtually all cellular processes, and a single mutation within a PPI can significantly impact protein functionality, potentially leading to diseases. While numerous approaches have emerged to predict changes in the free energy of binding due to mutations (\u0394\u0394Gbind), most lack precision. Recently, protein language models (PLMs) have shown powerful predictive capabilities by leveraging both sequence and structural data from protein complexes, yet they have not been optimized specifically for \u0394\u0394Gbind prediction.<\/jats:p>\n                  <\/jats:sec>\n                  <jats:sec>\n                    <jats:title>Results<\/jats:title>\n                    <jats:p>We developed an approach, ProBASS (Protein Binding Affinity from Structure and Sequence), to predict the effects of mutations on \u0394\u0394Gbind using two most advanced PLMs, ESM2 and ESM-IF1, which incorporate sequence and structural features, respectively. We first generated embeddings for each PPI mutant from the two PLMs and then fine-tuned ProBASS by training on a large dataset of experimental \u0394\u0394Gbind values. When training and testing were done on the same PPI, ProBASS achieved correlations with experimental \u0394\u0394Gbind values of 0.83\u2009\u00b1\u20090.05 and 0.69\u2009\u00b1\u20090.04 for single and double mutations, respectively. Additionally, when evaluated on a dataset of 2,325 single mutations across 131 PPIs, ProBASS reached a correlation of 0.81\u202f\u00b1\u202f0.02, substantially outperforming other PLMs in predictive accuracy. Our results demonstrate that refining pre-trained PLMs with extensive \u0394\u0394Gbind datasets across multiple PPIs is a successful approach for creating a precise and broadly applicable \u0394\u0394Gbind prediction model, facilitating future protein engineering and design studies. ProBASS\u2019s accuracy could be further improved through training as more experimental data becomes available.<\/jats:p>\n                  <\/jats:sec>\n                  <jats:sec>\n                    <jats:title>Availability and implementation<\/jats:title>\n                    <jats:p>ProBASS is available at: https:\/\/colab.research.google.com\/github\/sagagugit\/ProBASS\/blob\/main\/ProBASS.ipynb.<\/jats:p>\n                  <\/jats:sec>","DOI":"10.1093\/bioinformatics\/btaf270","type":"journal-article","created":{"date-parts":[[2025,5,8]],"date-time":"2025-05-08T07:33:08Z","timestamp":1746689588000},"source":"Crossref","is-referenced-by-count":4,"title":["ProBASS\u2014a language model with sequence and structural features for predicting the effect of mutations on binding affinity"],"prefix":"10.1093","volume":"41","author":[{"given":"Sagara N S","family":"Gurusinghe","sequence":"first","affiliation":[{"name":"Department of Biological Chemistry, The Alexander Silberman Institute of Life Sciences, The Hebrew University of Jerusalem , Jerusalem 9190401,","place":["Israel"]}]},{"given":"Yibing","family":"Wu","sequence":"additional","affiliation":[{"name":"Department of Pharmaceutical Chemistry, School of Pharmacy, University of California San Francisco , CA 94158,","place":["United States"]}]},{"given":"William","family":"DeGrado","sequence":"additional","affiliation":[{"name":"Department of Pharmaceutical Chemistry, School of Pharmacy, University of California San Francisco , CA 94158,","place":["United States"]}]},{"ORCID":"https:\/\/orcid.org\/0000-0001-6678-6497","authenticated-orcid":false,"given":"Julia M","family":"Shifman","sequence":"additional","affiliation":[{"name":"Department of Biological Chemistry, The Alexander Silberman Institute of Life Sciences, The Hebrew University of Jerusalem , Jerusalem 9190401,","place":["Israel"]}]}],"member":"286","published-online":{"date-parts":[[2025,5,9]]},"reference":[{"key":"2025061011520383700_btaf270-B1","doi-asserted-by":"crossref","first-page":"3","DOI":"10.1038\/nmeth0109-3","article-title":"Predicting free energy changes using structural ensembles","volume":"6","author":"Benedix","year":"2009","journal-title":"Nat Methods"},{"key":"2025061011520383700_btaf270-B2","doi-asserted-by":"crossref","first-page":"449","DOI":"10.1586\/14789450.2015.1084875","article-title":"How do oncoprotein mutations rewire protein-protein interaction networks?","volume":"12","author":"Bowler","year":"2015","journal-title":"Expert Rev Proteomics"},{"key":"2025061011520383700_btaf270-B3","doi-asserted-by":"crossref","first-page":"1478","DOI":"10.1002\/pmic.201100563","article-title":"History of protein-protein interactions: from egg-white to complex networks","volume":"12","author":"Braun","year":"2012","journal-title":"Proteomics"},{"key":"2025061011520383700_btaf270-B4","doi-asserted-by":"crossref","first-page":"e1005690","DOI":"10.1371\/journal.pcbi.1005690","article-title":"TopologyNet: topology based deep convolutional and multi-task neural networks for biomolecular property predictions","volume":"13","author":"Cang","year":"2017","journal-title":"PLoS Comput Biol"},{"key":"2025061011520383700_btaf270-B5","doi-asserted-by":"crossref","first-page":"S6","DOI":"10.1186\/1471-2164-14-S3-S6","article-title":"WS-SNPs&GO: a web server for predicting the deleterious effect of human protein variants using functional annotation","volume":"14","author":"Capriotti","year":"2013","journal-title":"BMC Genomics"},{"key":"2025061011520383700_btaf270-B6","doi-asserted-by":"crossref","first-page":"D829","DOI":"10.1093\/nar\/gkr929","article-title":"HotRegion: a database of predicted hot spot clusters","volume":"40","author":"Cukuroglu","year":"2012","journal-title":"Nucleic Acids Res"},{"key":"2025061011520383700_btaf270-B7","doi-asserted-by":"crossref","first-page":"49","DOI":"10.1126\/science.add2187","article-title":"Robust deep learning based protein sequence design using ProteinMPNN","volume":"378","author":"Dauparas","year":"2022","journal-title":"Science"},{"key":"2025061011520383700_btaf270-B8","doi-asserted-by":"crossref","first-page":"W333","DOI":"10.1093\/nar\/gkt450","article-title":"BeAtMuSiC: prediction of changes in protein\u2013protein binding affinity on mutations","volume":"41","author":"Dehouck","year":"2013","journal-title":"Nucleic Acids Res"},{"key":"2025061011520383700_btaf270-B9","doi-asserted-by":"crossref","first-page":"e2314853121","DOI":"10.1073\/pnas.2314853121","article-title":"Transfer learning to leverage larger datasets for improved prediction of protein stability changes","volume":"121","author":"Dieckhaus","year":"2024","journal-title":"Proc Natl Acad Sci U S A"},{"key":"2025061011520383700_btaf270-B10","doi-asserted-by":"crossref","first-page":"7112","DOI":"10.1109\/TPAMI.2021.3095381","article-title":"ProtTrans: toward understanding the language of life through Self-Supervised learning","volume":"44","author":"Elnaggar","year":"2022","journal-title":"IEEE Trans Pattern Anal Mach Intell"},{"key":"2025061011520383700_btaf270-B11","author":"Evans","year":"2022"},{"key":"2025061011520383700_btaf270-B12","doi-asserted-by":"crossref","first-page":"e1002819","DOI":"10.1371\/journal.pcbi.1002819","article-title":"Chapter 4: protein interactions and disease","volume":"8","author":"Gonzalez","year":"2012","journal-title":"PLoS Comput Biol"},{"key":"2025061011520383700_btaf270-B13","doi-asserted-by":"crossref","first-page":"369","DOI":"10.1016\/S0022-2836(02)00442-4","article-title":"Predicting changes in the stability of proteins and protein complexes: a study of more than 1000 mutations","volume":"320","author":"Guerois","year":"2002","journal-title":"J Mol Biol"},{"key":"2025061011520383700_btaf270-B14","doi-asserted-by":"crossref","first-page":"17261","DOI":"10.1021\/jacs.1c08707","article-title":"Climbing up and down binding landscapes through deep mutational scanning of three homologous Protein-Protein complexes","volume":"143","author":"Heyne","year":"2021","journal-title":"J Am Chem Soc"},{"key":"2025061011520383700_btaf270-B15","doi-asserted-by":"crossref","first-page":"297","DOI":"10.1038\/s41467-019-13895-8","article-title":"Generating quantitative binding landscapes through fractional binding selections combined with deep sequencing and data normalization","volume":"11","author":"Heyne","year":"2020","journal-title":"Nat Commun"},{"key":"2025061011520383700_btaf270-B16","doi-asserted-by":"crossref","first-page":"724297","DOI":"10.3389\/fbinf.2021.724297","article-title":"Detecting rewiring events in Protein-Protein interaction networks based on transcriptomic data","volume":"1","author":"Hollander","year":"2021","journal-title":"Front Bioinform"},{"key":"2025061011520383700_btaf270-B17","author":"Hsu"},{"key":"2025061011520383700_btaf270-B18","author":"Ingraham","year":"2019"},{"key":"2025061011520383700_btaf270-B19","doi-asserted-by":"crossref","first-page":"462","DOI":"10.1093\/bioinformatics\/bty635","article-title":"SKEMPI 2.0: an updated benchmark of changes in protein\u2013protein binding energy, kinetics and thermodynamics upon mutation","volume":"35","author":"Jankauskaite","year":"2019","journal-title":"Bioinformatics"},{"key":"2025061011520383700_btaf270-B20","doi-asserted-by":"crossref","first-page":"745","DOI":"10.1016\/j.jmb.2008.03.055","article-title":"Experimental and computational analyses of the energetic basis for dual recognition of immunity proteins by colicin endonucleases","volume":"379","author":"Keeble","year":"2008","journal-title":"J Mol Biol"},{"key":"2025061011520383700_btaf270-B21","doi-asserted-by":"crossref","first-page":"pl2","DOI":"10.1126\/stke.2192004pl2","article-title":"Computational alanine scanning of protein-protein interfaces","volume":"2004","author":"Kortemme","year":"2004","journal-title":"Sci STKE"},{"key":"2025061011520383700_btaf270-B22","doi-asserted-by":"crossref","first-page":"W480","DOI":"10.1093\/nar\/gkq471","article-title":"DrugScorePPI webserver: fast and accurate in silico alanine scanning for scoring protein-protein interactions","volume":"38","author":"Kr\u00fcger","year":"2010","journal-title":"Nucleic Acids Res"},{"key":"2025061011520383700_btaf270-B23","doi-asserted-by":"crossref","first-page":"145","DOI":"10.1142\/9789811211874_0006","volume-title":"Protein Interactions: Computational Methods, Analysis And Applications","author":"Kumar","year":"2020"},{"key":"2025061011520383700_btaf270-B24","doi-asserted-by":"crossref","first-page":"791","DOI":"10.1016\/S1074-5521(96)90063-0","article-title":"Sensing the heat: the application of isothermal titration calorimetry to thermodynamic studies of biomolecular interactions","volume":"3","author":"Ladbury","year":"1996","journal-title":"Chem Biol"},{"key":"2025061011520383700_btaf270-B25","doi-asserted-by":"crossref","first-page":"992","DOI":"10.1093\/bioinformatics\/btaa761","article-title":"SAAMBE-SEQ: a sequence-based method for predicting mutation effect on protein-protein binding affinity","volume":"37","author":"Li","year":"2021","journal-title":"Bioinformatics"},{"key":"2025061011520383700_btaf270-B26","doi-asserted-by":"crossref","first-page":"W494","DOI":"10.1093\/nar\/gkw374","article-title":"MutaBind estimates and interprets the effects of sequence variants on protein-protein interactions","volume":"44","author":"Li","year":"2016","journal-title":"Nucleic Acids Res"},{"key":"2025061011520383700_btaf270-B27","doi-asserted-by":"crossref","first-page":"1770","DOI":"10.1021\/ct401022c","article-title":"Predicting the impact of missense mutations on protein-protein binding affinity","volume":"10","author":"Li","year":"2014","journal-title":"J Chem Theory Comput"},{"key":"2025061011520383700_btaf270-B28","doi-asserted-by":"crossref","first-page":"11771","DOI":"10.1021\/bi9808621","article-title":"Dual recognition and the role of specificity-determining residues in colicin E9 DNase-immunity protein interactions","volume":"37","author":"Li","year":"1998","journal-title":"Biochemistry"},{"key":"2025061011520383700_btaf270-B29","author":"Lin","year":"2022"},{"key":"2025061011520383700_btaf270-B30","author":"Meier"},{"key":"2025061011520383700_btaf270-B31","doi-asserted-by":"crossref","first-page":"991","DOI":"10.1016\/S0022-2836(02)01281-0","article-title":"Structural characterization and functional significance of transient protein-protein interactions","volume":"325","author":"Nooren","year":"2003","journal-title":"J Mol Biol"},{"key":"2025061011520383700_btaf270-B32","doi-asserted-by":"crossref","first-page":"2563","DOI":"10.3390\/ijms21072563","article-title":"SAAMBE-3D: predicting effect of mutations on protein\u2013protein interactions","volume":"21","author":"Pahari","year":"2020","journal-title":"Int J Mol Sci"},{"key":"2025061011520383700_btaf270-B33","doi-asserted-by":"crossref","first-page":"e1004276","DOI":"10.1371\/journal.pcbi.1004276","article-title":"Predicting binding free energy change caused by point mutations with Knowledge-Modified MM\/PBSA method","volume":"11","author":"Petukh","year":"2015","journal-title":"PLoS Comput Bio"},{"key":"2025061011520383700_btaf270-B34","doi-asserted-by":"crossref","first-page":"547","DOI":"10.3390\/ijms17040547","article-title":"SAAMBE: webserver to predict the charge of binding free energy caused by amino acids mutations","volume":"17","author":"Petukh","year":"2016","journal-title":"Int J Mol Sci"},{"key":"2025061011520383700_btaf270-B35","doi-asserted-by":"crossref","first-page":"335","DOI":"10.1093\/bioinformatics\/btt691","article-title":"mCSM: predicting the effects of mutations in proteins using graph-based signatures","volume":"30","author":"Pires","year":"2014","journal-title":"Bioinformatics"},{"key":"2025061011520383700_btaf270-B36","author":"Prokhorenkova"},{"key":"2025061011520383700_btaf270-B37","author":"Rao"},{"key":"2025061011520383700_btaf270-B38","doi-asserted-by":"crossref","first-page":"167870","DOI":"10.1016\/j.jmb.2022.167870","article-title":"MPAD: a database for binding affinity of membrane protein-protein complexes and their mutants","volume":"435","author":"Ridha","year":"2023","journal-title":"J Mol Biol"},{"key":"2025061011520383700_btaf270-B39","doi-asserted-by":"crossref","first-page":"441","DOI":"10.1016\/j.sbi.2005.06.001","article-title":"Protein-protein interactions in human disease","volume":"15","author":"Ryan","year":"2005","journal-title":"Curr Opin Struct Biol"},{"key":"2025061011520383700_btaf270-B40","doi-asserted-by":"crossref","first-page":"1728","DOI":"10.1038\/s41467-022-29268-7","article-title":"Current progress and open challenges for applying deep learning across the biosciences","volume":"13","author":"Sapoval","year":"2022","journal-title":"Nat Commun"},{"key":"2025061011520383700_btaf270-B41","doi-asserted-by":"crossref","first-page":"1166","DOI":"10.1042\/BST20130121","article-title":"Predicting affinity- and specificity-enhancing mutations at protein-protein interfaces","volume":"41","author":"Sharabi","year":"2013","journal-title":"Biochem Soc Trans"},{"key":"2025061011520383700_btaf270-B42","doi-asserted-by":"crossref","first-page":"1295","DOI":"10.1016\/j.cell.2020.08.012","article-title":"Deep mutational scanning of SARS-CoV-2 receptor binding domain reveals constraints on folding and ACE2 binding","volume":"182","author":"Starr","year":"2020","journal-title":"Cell"},{"key":"2025061011520383700_btaf270-B43","doi-asserted-by":"crossref","first-page":"166810","DOI":"10.1016\/j.jmb.2021.166810","article-title":"ELASPIC2 (EL2): combining contextualized language models and graph neural networks to predict effects of mutations","volume":"433","author":"Strokach","year":"2021","journal-title":"J Mol Biol"},{"key":"2025061011520383700_btaf270-B44","doi-asserted-by":"crossref","first-page":"402","DOI":"10.1016\/j.cels.2020.08.016","article-title":"Fast and flexible protein design using deep graph neural networks","volume":"11","author":"Strokach","year":"2020","journal-title":"Cell Syst"},{"key":"2025061011520383700_btaf270-B45","doi-asserted-by":"crossref","first-page":"W402","DOI":"10.1093\/nar\/gkq323","article-title":"HotPoint: hot spot prediction server for protein interfaces","volume":"38","author":"Tuncbag","year":"2010","journal-title":"Nucleic Acids Res"},{"key":"2025061011520383700_btaf270-B46","author":"Vig"},{"key":"2025061011520383700_btaf270-B47","doi-asserted-by":"crossref","first-page":"201","DOI":"10.1007\/978-1-59745-483-4_14","article-title":"Analysis of biomolecules using surface plasmons","volume":"544","author":"Willander","year":"1994","journal-title":"Methods Mol Biol"},{"key":"2025061011520383700_btaf270-B48","doi-asserted-by":"crossref","first-page":"1589","DOI":"10.1093\/bioinformatics\/btw031","article-title":"ELASPIC web-server: proteome-wide structure-based prediction of mutation effects on protein stability and binding affinity","volume":"32","author":"Witvliet","year":"2016","journal-title":"Bioinformatics"},{"key":"2025061011520383700_btaf270-B49","doi-asserted-by":"crossref","first-page":"426","DOI":"10.1016\/j.jmb.2016.11.022","article-title":"BindProfX: assessing Mutation-Induced binding affinity change by protein interface profiles with Pseudo-Counts","volume":"429","author":"Xiong","year":"2017","journal-title":"J Mol Biol"},{"key":"2025061011520383700_btaf270-B50","doi-asserted-by":"crossref","first-page":"2692","DOI":"10.1016\/j.jmb.2014.04.026","article-title":"SuSPect: enhanced prediction of single amino acid variant (SAV) phenotype using network features","volume":"426","author":"Yates","year":"2014","journal-title":"J Mol Biol"},{"key":"2025061011520383700_btaf270-B51","doi-asserted-by":"crossref","first-page":"100939","DOI":"10.1016\/j.isci.2020.100939","article-title":"MutaBind2: predicting the impacts of single and multiple mutations on Protein-Protein interactions","volume":"23","author":"Zhang","year":"2020","journal-title":"iScience"},{"key":"2025061011520383700_btaf270-B52","doi-asserted-by":"crossref","first-page":"e1003592","DOI":"10.1371\/journal.pcbi.1003592","article-title":"Determining effects of non-synonymous SNPs on Protein-Protein interactions using supervised and semi-supervised learning","volume":"10","author":"Zhao","year":"2014","journal-title":"PLoS Comput Biol"},{"key":"2025061011520383700_btaf270-B53","author":"Zheng","year":"2024"},{"key":"2025061011520383700_btaf270-B54","doi-asserted-by":"crossref","first-page":"2671","DOI":"10.1002\/prot.23094","article-title":"KFC2: a knowledge-based hot spot prediction method based on interface solvation, atomic density, and plasticity features","volume":"79","author":"Zhu","year":"2011","journal-title":"Proteins: Structure, Function, and Bioinformatics"}],"container-title":["Bioinformatics"],"original-title":[],"language":"en","link":[{"URL":"https:\/\/academic.oup.com\/bioinformatics\/advance-article-pdf\/doi\/10.1093\/bioinformatics\/btaf270\/63136788\/btaf270.pdf","content-type":"application\/pdf","content-version":"am","intended-application":"syndication"},{"URL":"https:\/\/academic.oup.com\/bioinformatics\/article-pdf\/41\/5\/btaf270\/63136788\/btaf270.pdf","content-type":"application\/pdf","content-version":"vor","intended-application":"syndication"},{"URL":"https:\/\/academic.oup.com\/bioinformatics\/article-pdf\/41\/5\/btaf270\/63136788\/btaf270.pdf","content-type":"unspecified","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2025,6,10]],"date-time":"2025-06-10T11:52:21Z","timestamp":1749556341000},"score":1,"resource":{"primary":{"URL":"https:\/\/academic.oup.com\/bioinformatics\/article\/doi\/10.1093\/bioinformatics\/btaf270\/8127915"}},"subtitle":[],"editor":[{"given":"Arne","family":"Elofsson","sequence":"additional","affiliation":[]}],"short-title":[],"issued":{"date-parts":[[2025,5]]},"references-count":54,"journal-issue":{"issue":"5","published-print":{"date-parts":[[2025,5,6]]}},"URL":"https:\/\/doi.org\/10.1093\/bioinformatics\/btaf270","relation":{"has-preprint":[{"id-type":"doi","id":"10.1101\/2024.06.21.600041","asserted-by":"object"}]},"ISSN":["1367-4811"],"issn-type":[{"value":"1367-4811","type":"electronic"}],"subject":[],"published-other":{"date-parts":[[2025,5]]},"published":{"date-parts":[[2025,5]]},"article-number":"btaf270"}}