{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"institution":[{"name":"bioRxiv"}],"indexed":{"date-parts":[[2026,7,30]],"date-time":"2026-07-30T11:10:31Z","timestamp":1785409831315,"version":"3.56.0"},"posted":{"date-parts":[[2026,7,11]]},"group-title":"Biophysics","reference-count":78,"publisher":"openRxiv","license":[{"start":{"date-parts":[[2026,7,11]],"date-time":"2026-07-11T00:00:00Z","timestamp":1783728000000},"content-version":"vor","delay-in-days":0,"URL":"http:\/\/creativecommons.org\/licenses\/by-nc-nd\/4.0\/"}],"funder":[{"id":[{"id":"https:\/\/ror.org\/030689123","id-type":"ROR","asserted-by":"publisher"}]},{"award":["R01-GM148586"],"award-info":[{"award-number":["R01-GM148586"]}],"id":[{"id":"https:\/\/ror.org\/01cwqze88","id-type":"ROR","asserted-by":"publisher"}]}],"content-domain":{"domain":[],"crossmark-restriction":false},"short-container-title":[],"accepted":{"date-parts":[[2026,7,11]]},"abstract":"<jats:title>Abstract<\/jats:title>\n                <jats:p>\n                  Protein structure predictors achieve high single-state accuracy, but it remains unclear whether they can recover functionally relevant conformational ensembles or account for the presence of ligands and\/or binding partners. Here, we benchmark AlphaFold3, Boltz-2, Chai-1, and BioEmu on four canonical multi-state proteins (Pf-MATE, LAO, SecA, and \u03b2\n                  <jats:sub>2<\/jats:sub>\n                  AR), quantifying state bias and sampling breadth against experimental reference structures. Models frequently default to a dominant state represented in the PDB; small-molecule ligands have weak or inconsistent effects, while large protein partners drive clear conformational switching between states. Multiple sequence alignment (MSA)-based approaches (AF-Cluster and random subsampling) recapitulate similar biases, indicating that this behavior is not unique to newer architectures. These results underscore current limitations for multi-state protein structure prediction and structure-guided ligand discovery.\n                <\/jats:p>\n                <jats:sec>\n                  <jats:title>TOC Graphic<\/jats:title>\n                  <jats:fig id=\"ufig1\" position=\"float\" orientation=\"portrait\" fig-type=\"figure\">\n                    <jats:graphic xmlns:xlink=\"http:\/\/www.w3.org\/1999\/xlink\" xlink:href=\"737860v1_ufig1\" position=\"float\" orientation=\"portrait\"\/>\n                  <\/jats:fig>\n                <\/jats:sec>","DOI":"10.64898\/2026.07.10.737860","type":"posted-content","created":{"date-parts":[[2026,7,12]],"date-time":"2026-07-12T16:10:28Z","timestamp":1783872628000},"source":"Crossref","is-referenced-by-count":0,"title":["Benchmarking AI Protein Structure Predictors Reveals a Persistent Bias in Multi-State Proteins"],"prefix":"10.64898","author":[{"given":"Muhui","family":"Ye","sequence":"first","affiliation":[{"name":"School of Biological Sciences, Georgia Institute of Technology, Atlanta, GA 30332, USA"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Yu-Hong","family":"Wang","sequence":"additional","affiliation":[{"name":"School of Chemistry & Biochemistry, Georgia Institute of Technology, Atlanta, GA 30332, USA"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Maximilian","family":"Brogi","sequence":"additional","affiliation":[{"name":"School of Biological Sciences, Georgia Institute of Technology, Atlanta, GA 30332, USA"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Jerry M.","family":"Parks","sequence":"additional","affiliation":[{"name":"Biosciences Division, Oak Ridge National Laboratory, Oak Ridge, TN 37831, USA"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Katie M.","family":"Kuo","sequence":"additional","affiliation":[{"name":"School of Physics, Georgia Institute of Technology, Atlanta, GA 30332, USA"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"James C.","family":"Gumbart","sequence":"additional","affiliation":[{"name":"School of Physics, Georgia Institute of Technology, Atlanta, GA 30332, USA"},{"name":"School of Chemistry & Biochemistry, Georgia Institute of Technology, Atlanta, GA 30332, USA"}],"role":[{"vocabulary":"crossref","role":"author"}]}],"member":"54368","reference":[{"key":"2026071503450558000_2026.07.10.737860v1.1","doi-asserted-by":"publisher","DOI":"10.1038\/S41586-021-03819-2"},{"key":"2026071503450558000_2026.07.10.737860v1.2","doi-asserted-by":"crossref","first-page":"1607","DOI":"10.1002\/prot.26237","article-title":"Critical assessment of methods of protein structure prediction (CASP)\u2013Round XIV","volume":"89","year":"2021","journal-title":"Proteins: Struct., Funct., Bioinf"},{"key":"2026071503450558000_2026.07.10.737860v1.3","doi-asserted-by":"crossref","first-page":"1873","DOI":"10.1002\/prot.26382","article-title":"Multi-state modeling of G-protein coupled receptors at experimental accuracy","volume":"90","year":"2022","journal-title":"Proteins: Struct., Funct., Bioinf"},{"key":"2026071503450558000_2026.07.10.737860v1.4","doi-asserted-by":"publisher","DOI":"10.1038\/nature10349"},{"key":"2026071503450558000_2026.07.10.737860v1.5","doi-asserted-by":"publisher","DOI":"10.1073\/pnas.1209309109"},{"key":"2026071503450558000_2026.07.10.737860v1.6","doi-asserted-by":"publisher","DOI":"10.1016\/j.tibs.2010.04.009"},{"key":"2026071503450558000_2026.07.10.737860v1.7","doi-asserted-by":"publisher","DOI":"10.1093\/bioinformatics\/btac202"},{"key":"2026071503450558000_2026.07.10.737860v1.8","doi-asserted-by":"publisher","DOI":"10.7554\/eLife.75751"},{"key":"2026071503450558000_2026.07.10.737860v1.9","doi-asserted-by":"publisher","DOI":"10.1038\/S41586-024-07487-W"},{"key":"2026071503450558000_2026.07.10.737860v1.10","doi-asserted-by":"publisher","DOI":"10.48550\/arXiv.2510.15939"},{"key":"2026071503450558000_2026.07.10.737860v1.11","doi-asserted-by":"publisher","DOI":"10.1101\/2025.11.08.686881"},{"key":"2026071503450558000_2026.07.10.737860v1.12","doi-asserted-by":"publisher","DOI":"10.1101\/2024.10.10.615955"},{"key":"2026071503450558000_2026.07.10.737860v1.13","doi-asserted-by":"publisher","DOI":"10.1101\/2024.11.19.624167"},{"key":"2026071503450558000_2026.07.10.737860v1.14","doi-asserted-by":"publisher","DOI":"10.1101\/2025.06.14.659707"},{"key":"2026071503450558000_2026.07.10.737860v1.15","doi-asserted-by":"crossref","first-page":"eadv9817","DOI":"10.1126\/science.adv9817","article-title":"Scalable emulation of protein equilibrium ensembles with generative deep learning","volume":"389","year":"2025","journal-title":"Science"},{"key":"2026071503450558000_2026.07.10.737860v1.16","doi-asserted-by":"crossref","first-page":"11141","DOI":"10.1039\/D5CS00415B","article-title":"Modeling protein\u2013ligand interactions for drug discovery in the era of deep learning","volume":"54","year":"2025","journal-title":"Chem. Soc. Rev"},{"key":"2026071503450558000_2026.07.10.737860v1.17","doi-asserted-by":"publisher","DOI":"10.64898\/2025.12.04.692352"},{"key":"2026071503450558000_2026.07.10.737860v1.18","doi-asserted-by":"crossref","unstructured":"Kim, J. ; Correy, G. J. ; Hall, B. W. ; Rachman, M. M. ; Mailhot, O. ; Togo, T. ; Gonciarz, R. L. ; Jaishankar, P. ; Neitz, R. J. ; Hantz, E. R. ; Doruk, Y. U. ; Stevens, M. G. V. ; Diolaiti, M. E. ; Reid, R. ; Gopalkrishnan, S. ; Krogan, N. J. ; Renslo, A. R. ; Ashworth, A. ; Shoichet, B. K. ; Fraser, J. S . Large scale prospective evaluation of co-folding across 557 Mac1-ligand complexes and three virtual screens. Elife 2026, 15, RP110475.","DOI":"10.7554\/eLife.110475.1"},{"key":"2026071503450558000_2026.07.10.737860v1.19","doi-asserted-by":"crossref","first-page":"3671","DOI":"10.3390\/ijms26083671","article-title":"AlphaFold3: an overview of applications and performance insights","volume":"26","year":"2025","journal-title":"Int. J. Mol. Sci"},{"key":"2026071503450558000_2026.07.10.737860v1.20","doi-asserted-by":"publisher","DOI":"10.1038\/s41467-024-51801-z"},{"key":"2026071503450558000_2026.07.10.737860v1.21","doi-asserted-by":"publisher","DOI":"10.1016\/j.sbi.2024.102973"},{"key":"2026071503450558000_2026.07.10.737860v1.22","doi-asserted-by":"crossref","first-page":"103251","DOI":"10.1016\/j.sbi.2026.103251","article-title":"AI-based Methods for Simulating, Sampling, and Predicting Protein Ensembles","volume":"98","year":"2026","journal-title":"Curr. Opin. Struct. Biol"},{"key":"2026071503450558000_2026.07.10.737860v1.23","doi-asserted-by":"publisher","DOI":"10.1101\/2025.06.12.659353"},{"key":"2026071503450558000_2026.07.10.737860v1.24","first-page":"57","article-title":"Studying Ligand-Protein Interactions in the Era of Artificial Intelligence: Benchmarking Boltz-1 for 3D-Structure Prediction of Biomolecular Complexes","volume":"44","year":"2025","journal-title":"Kharkov Univ. Bull. Chem. Ser"},{"key":"2026071503450558000_2026.07.10.737860v1.25","doi-asserted-by":"publisher","DOI":"10.1038\/nature12014"},{"key":"2026071503450558000_2026.07.10.737860v1.26","doi-asserted-by":"publisher","DOI":"10.1016\/S0021-9258(18)82131-7"},{"key":"2026071503450558000_2026.07.10.737860v1.27","doi-asserted-by":"publisher","DOI":"10.1016\/S0021-9258(17)41754-6"},{"key":"2026071503450558000_2026.07.10.737860v1.28","doi-asserted-by":"publisher","DOI":"10.1126\/science.1074424"},{"key":"2026071503450558000_2026.07.10.737860v1.29","doi-asserted-by":"publisher","DOI":"10.1038\/nature07335"},{"key":"2026071503450558000_2026.07.10.737860v1.30","doi-asserted-by":"publisher","DOI":"10.1073\/pnas.0811065106"},{"key":"2026071503450558000_2026.07.10.737860v1.31","doi-asserted-by":"publisher","DOI":"10.1016\/j.bbapap.2008.11.012"},{"key":"2026071503450558000_2026.07.10.737860v1.32","doi-asserted-by":"publisher","DOI":"10.1016\/j.tips.2006.09.001"},{"key":"2026071503450558000_2026.07.10.737860v1.33","doi-asserted-by":"publisher","DOI":"10.1038\/nsmb.2687"},{"key":"2026071503450558000_2026.07.10.737860v1.34","doi-asserted-by":"publisher","DOI":"10.1073\/pnas.1904210116"},{"key":"2026071503450558000_2026.07.10.737860v1.35","doi-asserted-by":"crossref","first-page":"26323","DOI":"10.1016\/S0021-9258(18)47196-7","article-title":"Structural basis for multiple ligand specificity of the periplasmic lysine-, arginine-, ornithine-binding protein","volume":"269","year":"1994","journal-title":"J. Biol. Chem"},{"key":"2026071503450558000_2026.07.10.737860v1.36","doi-asserted-by":"crossref","unstructured":"Felder, C. B. ; Graul, R. C. ; Lee, A. Y. ; Merkle, H.-P. ; Sadee, W . The Venus flytrap of periplasmic binding proteins: an ancient protein module present in multiple drug receptors. AAPS PharmSci. 1999, 1, E2.","DOI":"10.1208\/ps010202"},{"key":"2026071503450558000_2026.07.10.737860v1.37","doi-asserted-by":"crossref","first-page":"1466","DOI":"10.1016\/j.bbamcr.2014.02.014","article-title":"SecA-mediated targeting and translocation of secretory proteins","volume":"1843","year":"2014","journal-title":"Biochim. Biophys. Acta - Mol. Cell Res"},{"key":"2026071503450558000_2026.07.10.737860v1.38","doi-asserted-by":"publisher","DOI":"10.1038\/nrmicro.2016.161"},{"key":"2026071503450558000_2026.07.10.737860v1.39","doi-asserted-by":"publisher","DOI":"10.1073\/pnas.0910550106"},{"key":"2026071503450558000_2026.07.10.737860v1.40","doi-asserted-by":"publisher","DOI":"10.1016\/j.cell.2014.03.063"},{"key":"2026071503450558000_2026.07.10.737860v1.41","doi-asserted-by":"publisher","DOI":"10.1016\/j.jmb.2015.05.003"},{"key":"2026071503450558000_2026.07.10.737860v1.42","doi-asserted-by":"crossref","first-page":"130","DOI":"10.1038\/s42003-018-0133-4","article-title":"Large conformational changes of a highly dynamic pre-protein binding domain in SecA","volume":"1","year":"2018","journal-title":"Commun. Biol"},{"key":"2026071503450558000_2026.07.10.737860v1.43","doi-asserted-by":"publisher","DOI":"10.1073\/pnas.0401742101"},{"key":"2026071503450558000_2026.07.10.737860v1.44","doi-asserted-by":"publisher","DOI":"10.1016\/j.cell.2025.01.037"},{"key":"2026071503450558000_2026.07.10.737860v1.45","doi-asserted-by":"publisher","DOI":"10.1016\/j.ando.2020.03.012"},{"key":"2026071503450558000_2026.07.10.737860v1.46","doi-asserted-by":"crossref","first-page":"E24","DOI":"10.2527\/animalsci2002.0021881200800ES10005x","article-title":"\u03b2-adrenergic receptor modulation of adipocyte metabolism and growth","volume":"80","year":"2002","journal-title":"J. Anim. Sci"},{"key":"2026071503450558000_2026.07.10.737860v1.47","doi-asserted-by":"crossref","first-page":"105","DOI":"10.14348\/molcells.2015.2301","article-title":"Structural features of \u03b22 adrenergic receptor: Crystal structures and beyond","volume":"38","year":"2015","journal-title":"Mol. Cells"},{"key":"2026071503450558000_2026.07.10.737860v1.48","doi-asserted-by":"publisher","DOI":"10.1038\/nature10361"},{"key":"2026071503450558000_2026.07.10.737860v1.49","doi-asserted-by":"publisher","DOI":"10.1073\/pnas.1519626112"},{"key":"2026071503450558000_2026.07.10.737860v1.50","doi-asserted-by":"publisher","DOI":"10.1016\/j.cell.2013.01.008"},{"key":"2026071503450558000_2026.07.10.737860v1.51","doi-asserted-by":"publisher","DOI":"10.1038\/s41586-023-06832-9"},{"key":"2026071503450558000_2026.07.10.737860v1.52","doi-asserted-by":"crossref","unstructured":"Mitjavila-Domen\u00e8ch, N. ; D\u00edaz-Holgu\u00edn, A. ; Hu, H. ; Kahlous, N. A. ; Cabeza de Vaca, I. ; Wallner, B. ; Carlsson, J. Improving AlphaFold2 performance in virtual screens targeting GPCRs by enhancing binding-site conformational sampling. J. Chem. Inf. Model. 2026, In press.","DOI":"10.1021\/acs.jcim.6c00034"},{"key":"2026071503450558000_2026.07.10.737860v1.53","doi-asserted-by":"publisher","DOI":"10.1371\/journal.pcbi.1010483"},{"key":"2026071503450558000_2026.07.10.737860v1.54","doi-asserted-by":"publisher","DOI":"10.1038\/S41592-022-01488-1"},{"key":"2026071503450558000_2026.07.10.737860v1.55","doi-asserted-by":"publisher","DOI":"10.1038\/s41467-024-51507-2"},{"key":"2026071503450558000_2026.07.10.737860v1.56","doi-asserted-by":"publisher","DOI":"10.1021\/acs.jctc.3c00290"},{"key":"2026071503450558000_2026.07.10.737860v1.57","doi-asserted-by":"crossref","unstructured":"Gu, X. ; Aranganathan, A. ; Tiwary, P . Empowering AlphaFold2 for protein conformation selective drug discovery with AlphaFold2-RAVE. Elife 2024, 13, RP99702.","DOI":"10.7554\/eLife.99702"},{"key":"2026071503450558000_2026.07.10.737860v1.58","doi-asserted-by":"crossref","first-page":"12055","DOI":"10.1039\/D5SC08253F","article-title":"DeepPath: overcoming data scarcity for protein transition pathway prediction using physics-based deep learning","volume":"17","year":"2026","journal-title":"Chem. Sci"},{"key":"2026071503450558000_2026.07.10.737860v1.59","doi-asserted-by":"publisher","DOI":"10.1038\/s41589-024-01638-w"},{"key":"2026071503450558000_2026.07.10.737860v1.60","doi-asserted-by":"crossref","first-page":"1539","DOI":"10.1002\/prot.26617","article-title":"Critical assessment of methods of protein structure prediction (CASP)\u2014Round XV","volume":"91","year":"2023","journal-title":"Proteins: Struct., Funct., Bioinf"},{"key":"2026071503450558000_2026.07.10.737860v1.61","doi-asserted-by":"crossref","first-page":"330","DOI":"10.1002\/prot.70065","article-title":"Modeling alternative conformational states in CASP16","volume":"94","year":"2026","journal-title":"Proteins: Struct., Funct., Bioinf"},{"key":"2026071503450558000_2026.07.10.737860v1.62","doi-asserted-by":"crossref","first-page":"5","DOI":"10.1002\/prot.70076","article-title":"Progress and bottle-necks for deep learning in computational structure biology: CASP Round XVI","volume":"94","year":"2026","journal-title":"Proteins: Struct., Funct., Bioinf"},{"key":"2026071503450558000_2026.07.10.737860v1.63","doi-asserted-by":"crossref","first-page":"10167","DOI":"10.1021\/acs.jcim.5c01765","article-title":"Sampling and Ranking of Protein Conformations Using Machine Learning Techniques Do Not Improve the Quality of Rigid Protein\u2013Protein Docking","volume":"65","year":"2025","journal-title":"J. Chem. Inf. Model"},{"key":"2026071503450558000_2026.07.10.737860v1.64","doi-asserted-by":"crossref","first-page":"627","DOI":"10.1038\/s42003-026-10112-3","article-title":"The latest AI breakthroughs in structural biology: protein binder design and conformational state prediction","volume":"9","year":"2026","journal-title":"Commun. Biol"},{"key":"2026071503450558000_2026.07.10.737860v1.65","doi-asserted-by":"publisher","DOI":"10.1016\/j.sbi.2023.102645"},{"key":"2026071503450558000_2026.07.10.737860v1.66","doi-asserted-by":"publisher","DOI":"10.1038\/s41401-024-01429-y"},{"key":"2026071503450558000_2026.07.10.737860v1.67","doi-asserted-by":"crossref","first-page":"2896","DOI":"10.3390\/ijms27062896","article-title":"Assessing the Performance of BioEmu in Understanding Protein Dynamics","volume":"27","year":"2026","journal-title":"Int. J. Mol. Sci"},{"key":"2026071503450558000_2026.07.10.737860v1.68","doi-asserted-by":"crossref","first-page":"7168","DOI":"10.1021\/acs.jcim.6c01000","article-title":"Accelerated Sampling of Protein Dynamics Using BioEmu-Augmented Molecular Simulation","volume":"66","year":"2026","journal-title":"J. Chem. Inf. Model"},{"key":"2026071503450558000_2026.07.10.737860v1.69","doi-asserted-by":"publisher","DOI":"10.1073\/pnas.2530709123"},{"key":"2026071503450558000_2026.07.10.737860v1.70","doi-asserted-by":"crossref","first-page":"3355","DOI":"10.1038\/s41401-025-01617-4","article-title":"An update for Al-phaFold3 versus experimental structures: assessing the precision of small molecule binding in GPCRs","volume":"46","year":"2025","journal-title":"Acta Pharmacol. Sin"},{"key":"2026071503450558000_2026.07.10.737860v1.71","doi-asserted-by":"publisher","DOI":"10.1039\/D5SC06481C"},{"key":"2026071503450558000_2026.07.10.737860v1.72","doi-asserted-by":"publisher","DOI":"10.26434\/chemrxiv-2026-1rkqz"},{"key":"2026071503450558000_2026.07.10.737860v1.73","doi-asserted-by":"publisher","DOI":"10.1093\/nar\/gkae1010"},{"key":"2026071503450558000_2026.07.10.737860v1.74","doi-asserted-by":"publisher","DOI":"10.1093\/nar\/28.1.235"},{"key":"2026071503450558000_2026.07.10.737860v1.75","doi-asserted-by":"crossref","unstructured":"Wayment-Steele, H. K. ; Ovchinnikov, S. ; Colwell, L. ; Kern, D . Does Sequence Clustering Confound AlphaFold2? J. Mol. Biol. 2025, 437, 169376.","DOI":"10.1016\/j.jmb.2025.169376"},{"key":"2026071503450558000_2026.07.10.737860v1.76","unstructured":"PyMOL The PyMOL Molecular Graphics System, Version 3.1.6.1, Schr\u00f6dinger, LLC."},{"key":"2026071503450558000_2026.07.10.737860v1.77","doi-asserted-by":"publisher","DOI":"10.1093\/nar\/gki524"},{"key":"2026071503450558000_2026.07.10.737860v1.78","doi-asserted-by":"publisher","DOI":"10.1093\/bioinformatics\/btq066"}],"container-title":[],"original-title":[],"link":[{"URL":"https:\/\/syndication.highwire.org\/content\/doi\/10.64898\/2026.07.10.737860","content-type":"unspecified","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2026,7,15]],"date-time":"2026-07-15T10:45:20Z","timestamp":1784112320000},"score":1,"resource":{"primary":{"URL":"http:\/\/biorxiv.org\/lookup\/doi\/10.64898\/2026.07.10.737860"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2026,7,11]]},"references-count":78,"URL":"https:\/\/doi.org\/10.64898\/2026.07.10.737860","relation":{},"subject":[],"published":{"date-parts":[[2026,7,11]]},"subtype":"preprint"}}