{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2026,7,13]],"date-time":"2026-07-13T19:58:22Z","timestamp":1783972702850,"version":"3.55.0"},"reference-count":29,"publisher":"Oxford University Press (OUP)","issue":"4","license":[{"start":{"date-parts":[[2026,3,21]],"date-time":"2026-03-21T00:00:00Z","timestamp":1774051200000},"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":[[2026,4,7]]},"abstract":"<jats:title>Abstract<\/jats:title>\n                  <jats:sec>\n                    <jats:title>Motivation<\/jats:title>\n                    <jats:p>AlphaFold2 significantly improved the prediction of protein complex structures. However, its accuracy is lower for interactions without coevolutionary signals, such as host\u2013pathogen and antibody\u2013antigen interactions. Two strategies have been developed to address this limitation: massive sampling and replacing the evoformer with the pairformer, which does not rely on coevolution, as introduced in AlphaFold3, thereby enabling more structural reasoning by the network.<\/jats:p>\n                  <\/jats:sec>\n                  <jats:sec>\n                    <jats:title>Results<\/jats:title>\n                    <jats:p>In this study, we benchmark structure prediction methods on unseen antibody\u2013antigen complexes. We found that increased sampling improves the chances of generating a correct protein model, roughly in a log-linear manner. However, the internal quality estimates by AlphaFold often cannot identify the best predicted structures for each target, resulting in a significant loss of performance for the top-ranked protein model compared with the best model. For all methods, a significant challenge remains the identification of the best model. We also show that AlphaFold3 outperforms AlphaFold2, Boltz-1, and Chai-1. Furthermore, AlphaFold3 performance declines significantly for complexes that lack structural similarity to the training set, indicating that it has to some extent learned to detect remote structural similarities.<\/jats:p>\n                  <\/jats:sec>\n                  <jats:sec>\n                    <jats:title>Availability and implementation<\/jats:title>\n                    <jats:p>All code is available from github.com\/samuelfromm\/abag-benchmark-set\/ and all data from DOI: 10.5281\/zenodo.17978681. The latter repository also contains the code.<\/jats:p>\n                  <\/jats:sec>","DOI":"10.1093\/bioinformatics\/btag136","type":"journal-article","created":{"date-parts":[[2026,3,21]],"date-time":"2026-03-21T10:23:02Z","timestamp":1774088582000},"source":"Crossref","is-referenced-by-count":6,"title":["Evaluating deep learning based structure prediction methods on antibody\u2013antigen complexes"],"prefix":"10.1093","volume":"42","author":[{"given":"Samuel","family":"Fromm","sequence":"first","affiliation":[{"name":"Science for Life Laboratory and Department of Biochemistry and Biophysics, Stockholm University , Solna 171 21,","place":["Sweden"]}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Marko","family":"Ludaic","sequence":"additional","affiliation":[{"name":"Science for Life Laboratory and Department of Biochemistry and Biophysics, Stockholm University , Solna 171 21,","place":["Sweden"]}],"role":[{"vocabulary":"crossref","role":"author"}]},{"ORCID":"https:\/\/orcid.org\/0000-0002-7115-9751","authenticated-orcid":false,"given":"Arne","family":"Elofsson","sequence":"additional","affiliation":[{"name":"Science for Life Laboratory and Department of Biochemistry and Biophysics, Stockholm University , Solna 171 21,","place":["Sweden"]}],"role":[{"vocabulary":"crossref","role":"author"}]}],"member":"286","published-online":{"date-parts":[[2026,3,21]]},"reference":[{"key":"2026042409462071200_btag136-B1","doi-asserted-by":"crossref","first-page":"E4","DOI":"10.1038\/s41586-024-08416-7","article-title":"Addendum: accurate structure prediction of biomolecular interactions with AlphaFold 3","volume":"636","author":"Abramson","year":"2024","journal-title":"Nature"},{"key":"2026042409462071200_btag136-B2","doi-asserted-by":"crossref","first-page":"e0161879","DOI":"10.1371\/journal.pone.0161879","article-title":"DockQ: a quality measure for protein\u2013protein docking models","volume":"11","author":"Basu","year":"2016","journal-title":"PLoS One"},{"key":"2026042409462071200_btag136-B3","doi-asserted-by":"crossref","first-page":"1265","DOI":"10.1038\/s41467-022-28865-w","article-title":"Improved prediction of protein\u2013protein interactions using AlphaFold2","volume":"13","author":"Bryant","year":"2022","journal-title":"Nat Commun"},{"key":"2026042409462071200_btag136-B4","doi-asserted-by":"crossref","DOI":"10.7554\/eLife.75751","article-title":"Sampling alternative conformational states of transporters and receptors with AlphaFold2","volume":"11","author":"Del Alamo","year":"2022","journal-title":"Elife"},{"key":"2026042409462071200_btag136-B5","author":"Discovery","year":"2024"},{"key":"2026042409462071200_btag136-B6","author":"Dunbrack","year":"2025"},{"key":"2026042409462071200_btag136-B7","author":"Elofsson","year":"2025"},{"key":"2026042409462071200_btag136-B8","author":"Evans","year":"2022"},{"key":"2026042409462071200_btag136-B9","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":"2026042409462071200_btag136-B10","doi-asserted-by":"crossref","first-page":"373","DOI":"10.1038\/s42003-025-07791-9","article-title":"AFsample2 predicts multiple conformations and ensembles with AlphaFold2","volume":"8","author":"Kalakoti","year":"2025","journal-title":"Commun Biol"},{"key":"2026042409462071200_btag136-B11","doi-asserted-by":"crossref","first-page":"469","DOI":"10.1038\/s41592-025-02593-7","article-title":"Rapid and sensitive protein complex alignment with foldseek-multimer","volume":"22","author":"Kim","year":"2025","journal-title":"Nat Methods"},{"key":"2026042409462071200_btag136-B12","doi-asserted-by":"crossref","first-page":"1","DOI":"10.1186\/s12929-019-0592-z","article-title":"Development of therapeutic antibodies for the treatment of diseases","volume":"27","author":"Lu","year":"2020","journal-title":"J Biomed Sci"},{"key":"2026042409462071200_btag136-B13","doi-asserted-by":"crossref","first-page":"e5127","DOI":"10.1002\/pro.5127","article-title":"A comparison of antibody\u2013antigen complex sequence-to-structure prediction methods and their systematic biases","volume":"33","author":"McCoy","year":"2024","journal-title":"Protein Sci"},{"key":"2026042409462071200_btag136-B14","doi-asserted-by":"crossref","DOI":"10.1093\/bioinformatics\/btae586","article-title":"DockQ v2: improved automatic quality measure for protein multimers, nucleic acids, and small molecules","volume":"40","author":"Mirabello","year":"2024","journal-title":"Bioinformatics"},{"key":"2026042409462071200_btag136-B15","doi-asserted-by":"crossref","first-page":"679","DOI":"10.1038\/s41592-022-01488-1","article-title":"Colabfold: making protein folding accessible to all","volume":"19","author":"Mirdita","year":"2022","journal-title":"Nat Methods"},{"key":"2026042409462071200_btag136-B16","doi-asserted-by":"crossref","first-page":"e83","DOI":"10.1093\/nar\/gkp318","article-title":"MM-align: a quick algorithm for aligning multiple-chain protein complex structures using iterative dynamic programming","volume":"37","author":"Mukherjee","year":"2009","journal-title":"Nucleic Acids Res"},{"key":"2026042409462071200_btag136-B17","doi-asserted-by":"crossref","first-page":"954","DOI":"10.1093\/bioinformatics\/btab760","article-title":"Limits and potential of combined folding and docking","volume":"38","author":"Pozzati","year":"2022","journal-title":"Bioinformatics"},{"key":"2026042409462071200_btag136-B18","article-title":"Massive sampling strategy for antibody\u2013antigen targets in CAPRI round 55 with MassiveFold","author":"Raouraoua","year":"2026","journal-title":"Proteins"},{"key":"2026042409462071200_btag136-B19","doi-asserted-by":"crossref","first-page":"824","DOI":"10.1038\/s43588-024-00714-4","article-title":"MassiveFold: unveiling AlphaFold\u2019s hidden potential with optimized and parallelized massive sampling","volume":"4","author":"Raouraoua","year":"2024","journal-title":"Nat Comput Sci"},{"key":"2026042409462071200_btag136-B20","doi-asserted-by":"crossref","first-page":"e1013168","DOI":"10.1371\/journal.pcbi.1013168","article-title":"AI-first structural identification of pathogenic protein target interfaces","volume":"21","author":"Saluri","year":"2025","journal-title":"PLoS Comput Biol"},{"key":"2026042409462071200_btag136-B21","doi-asserted-by":"crossref","first-page":"D1368","DOI":"10.1093\/nar\/gkab1050","article-title":"SAbDab in the age of biotherapeutics: updates including SAbDab-nano, the nanobody structure tracker","volume":"50","author":"Schneider","year":"2022","journal-title":"Nucleic Acids Res"},{"key":"2026042409462071200_btag136-B22","doi-asserted-by":"crossref","first-page":"243","DOI":"10.1038\/s41587-023-01773-0","article-title":"Fast and accurate protein structure search with foldseek","volume":"42","author":"van Kempen","year":"2024","journal-title":"Nat Biotechnol"},{"key":"2026042409462071200_btag136-B23","volume-title":"Bioinformatics","author":"Varga"},{"key":"2026042409462071200_btag136-B24","doi-asserted-by":"crossref","DOI":"10.1093\/bioinformatics\/btad573","article-title":"AFsample: improving multimer prediction with AlphaFold using massive sampling","volume":"39","author":"Wallner","year":"2023","journal-title":"Bioinformatics"},{"key":"2026042409462071200_btag136-B25","author":"Wohlwend","year":"2025"},{"key":"2026042409462071200_btag136-B26","doi-asserted-by":"crossref","first-page":"e4865","DOI":"10.1002\/pro.4865","article-title":"Evaluation of AlphaFold antibody\u2013antigen modeling with implications for improving predictive accuracy","volume":"33","author":"Yin","year":"2024","journal-title":"Protein Sci"},{"key":"2026042409462071200_btag136-B27","doi-asserted-by":"crossref","first-page":"1109","DOI":"10.1038\/s41592-022-01585-1","article-title":"US-align: universal structure alignments of proteins, nucleic acids, and macromolecular complexes","volume":"19","author":"Zhang","year":"2022","journal-title":"Nat Methods"},{"key":"2026042409462071200_btag136-B28","doi-asserted-by":"crossref","first-page":"702","DOI":"10.1002\/prot.20264","article-title":"Scoring function for automated assessment of protein structure template quality","volume":"57","author":"Zhang","year":"2004","journal-title":"Proteins"},{"key":"2026042409462071200_btag136-B29","doi-asserted-by":"crossref","first-page":"btad424","DOI":"10.1093\/bioinformatics\/btad424","article-title":"Evaluation of AlphaFold-multimer prediction on multi-chain protein complexes","volume":"39","author":"Zhu","year":"2023","journal-title":"Bioinformatics"}],"container-title":["Bioinformatics"],"original-title":[],"language":"en","link":[{"URL":"https:\/\/academic.oup.com\/bioinformatics\/advance-article-pdf\/doi\/10.1093\/bioinformatics\/btag136\/67456922\/btag136.pdf","content-type":"application\/pdf","content-version":"am","intended-application":"syndication"},{"URL":"https:\/\/academic.oup.com\/bioinformatics\/article-pdf\/42\/4\/btag136\/67456922\/btag136.pdf","content-type":"application\/pdf","content-version":"vor","intended-application":"syndication"},{"URL":"https:\/\/academic.oup.com\/bioinformatics\/article-pdf\/42\/4\/btag136\/67456922\/btag136.pdf","content-type":"unspecified","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2026,4,24]],"date-time":"2026-04-24T13:46:33Z","timestamp":1777038393000},"score":1,"resource":{"primary":{"URL":"https:\/\/academic.oup.com\/bioinformatics\/article\/doi\/10.1093\/bioinformatics\/btag136\/8533244"}},"subtitle":[],"editor":[{"given":"Pier Luigi","family":"Martelli","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"editor"}]}],"short-title":[],"issued":{"date-parts":[[2026,3,21]]},"references-count":29,"journal-issue":{"issue":"4","published-print":{"date-parts":[[2026,4,7]]}},"URL":"https:\/\/doi.org\/10.1093\/bioinformatics\/btag136","relation":{},"ISSN":["1367-4803","1367-4811"],"issn-type":[{"value":"1367-4803","type":"print"},{"value":"1367-4811","type":"electronic"}],"subject":[],"published-other":{"date-parts":[[2026,4]]},"published":{"date-parts":[[2026,3,21]]},"article-number":"btag136"}}