{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"institution":[{"name":"bioRxiv"}],"indexed":{"date-parts":[[2026,9,1]],"date-time":"2026-09-01T18:57:46Z","timestamp":1788289066559,"version":"build-2803163510"},"posted":{"date-parts":[[2025,2,21]]},"group-title":"Biophysics","reference-count":69,"publisher":"openRxiv","license":[{"start":{"date-parts":[[2025,2,21]],"date-time":"2025-02-21T00:00:00Z","timestamp":1740096000000},"content-version":"vor","delay-in-days":0,"URL":"http:\/\/creativecommons.org\/licenses\/by\/4.0\/"}],"content-domain":{"domain":[],"crossmark-restriction":false},"short-container-title":[],"accepted":{"date-parts":[[2025,3,11]]},"abstract":"<jats:title>Abstract<\/jats:title>\n                <jats:p>Advances in machine learning have transformed structural biology, enabling swift and accurate prediction of protein structure from sequence. However, challenges persist in capturing sidechain packing, condition-dependent conformational dynamics, and biomolecular interactions, primarily due to scarcity of high-quality training data. Emerging techniques, including cryo-electron tomography (cryo-ET) and high-throughput crystallography, promise vast new sources of structural data, but translating experimental observations into mechanistically interpretable atomic models remains a key bottleneck. Here, we address these challenges by improving the efficiency of structural analysis through combining experimental measurements with a landmark protein structure prediction method \u2013 AlphaFold2. We present an augmentation of AlphaFold2, ROCKET, that refines its predictions using cryo-EM, cryo-ET, and X-ray crystallography data, and demonstrate that this approach captures biologically important structural variation that AlphaFold2 does not. By performing structure optimization in the space of coevolutionary embeddings, rather than Cartesian coordinates, ROCKET automates difficult modeling tasks, such as flips of functional loops and domain rearrangements, that are beyond the scope of current state-of-the-art methods and, in some instances, even manual human modeling. The ability to efficiently sample these barrier-crossing rearrangements unlocks a new horizon for scalable and automated model building. Crucially, ROCKET does not require retraining of AlphaFold2 and is readily adaptable to multimers, ligand-cofolding, and other data modalities. Conversely, our differentiable crystal-lographic and cryo-EM target functions are capable of augmenting other structure prediction methods. ROCKET thus provides an extensible framework for the integration of experimental observables with biomolecular machine learning.<\/jats:p>","DOI":"10.1101\/2025.02.18.638828","type":"posted-content","created":{"date-parts":[[2025,2,21]],"date-time":"2025-02-21T22:05:14Z","timestamp":1740175514000},"source":"Crossref","is-referenced-by-count":15,"title":["AlphaFold as a Prior: Experimental Structure Determination Conditioned on a Pretrained Neural Network"],"prefix":"10.64898","author":[{"ORCID":"https:\/\/orcid.org\/0000-0001-5246-9124","authenticated-orcid":false,"given":"Alisia","family":"Fadini","sequence":"first","affiliation":[{"name":"Cambridge Institute for Medical Research, University of Cambridge"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"ORCID":"https:\/\/orcid.org\/0000-0003-0560-7533","authenticated-orcid":false,"given":"Minhuan","family":"Li","sequence":"additional","affiliation":[{"name":"John A. Paulson School of Engineering & Applied Sciences, Harvard University"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"ORCID":"https:\/\/orcid.org\/0009-0001-6855-9631","authenticated-orcid":false,"given":"Airlie J.","family":"McCoy","sequence":"additional","affiliation":[{"name":"Cambridge Institute for Medical Research, University of Cambridge"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"ORCID":"https:\/\/orcid.org\/0000-0001-6384-0320","authenticated-orcid":false,"given":"Thomas C.","family":"Terwilliger","sequence":"additional","affiliation":[{"name":"New Mexico Consortium"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"ORCID":"https:\/\/orcid.org\/0000-0001-8273-0047","authenticated-orcid":false,"given":"Randy J.","family":"Read","sequence":"additional","affiliation":[{"name":"Cambridge Institute for Medical Research, University of Cambridge"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"ORCID":"https:\/\/orcid.org\/0000-0003-2332-9223","authenticated-orcid":false,"given":"Doeke","family":"Hekstra","sequence":"additional","affiliation":[{"name":"John A. Paulson School of Engineering & Applied Sciences, Harvard University"},{"name":"Department of Molecular and Cellular Biology, Harvard University"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"ORCID":"https:\/\/orcid.org\/0000-0001-6817-1322","authenticated-orcid":false,"given":"Mohammed","family":"AlQuraishi","sequence":"additional","affiliation":[{"name":"Department of Systems Biology, Columbia University"}],"role":[{"vocabulary":"crossref","role":"author"}]}],"member":"54368","reference":[{"key":"2026041100133766000_2025.02.18.638828v2.1","doi-asserted-by":"publisher","DOI":"10.1038\/S41586-021-03819-2"},{"key":"2026041100133766000_2025.02.18.638828v2.2","doi-asserted-by":"publisher","DOI":"10.1126\/science.abj8754"},{"key":"2026041100133766000_2025.02.18.638828v2.3","doi-asserted-by":"publisher","DOI":"10.1002\/prot.26257"},{"key":"2026041100133766000_2025.02.18.638828v2.4","doi-asserted-by":"publisher","DOI":"10.1002\/prot.26171"},{"key":"2026041100133766000_2025.02.18.638828v2.5","doi-asserted-by":"publisher","DOI":"10.7554\/eLife.89386.2"},{"key":"2026041100133766000_2025.02.18.638828v2.6","doi-asserted-by":"publisher","DOI":"10.1038\/s41592-023-02087-4"},{"key":"2026041100133766000_2025.02.18.638828v2.7","doi-asserted-by":"crossref","unstructured":"He, X.-h. , Li, J.-r. , Shen, S.-y. & Xu, H. E. AlphaFold3 versus experimental structures: assessment of the accuracy in ligand-bound G protein-coupled receptors. Acta Pharmacologica Sinica 1\u201312 (2024).","DOI":"10.1038\/s41401-024-01429-y"},{"key":"2026041100133766000_2025.02.18.638828v2.8","doi-asserted-by":"publisher","DOI":"10.1039\/D3SC04185A"},{"key":"2026041100133766000_2025.02.18.638828v2.9","doi-asserted-by":"publisher","DOI":"10.1038\/S41586-024-07487-W"},{"key":"2026041100133766000_2025.02.18.638828v2.10","doi-asserted-by":"publisher","DOI":"10.1126\/science.aba0954"},{"key":"2026041100133766000_2025.02.18.638828v2.11","doi-asserted-by":"publisher","DOI":"10.1146\/annurev-biophys-111622-091155"},{"key":"2026041100133766000_2025.02.18.638828v2.12","doi-asserted-by":"publisher","DOI":"10.1002\/1873-3468.13948"},{"key":"2026041100133766000_2025.02.18.638828v2.13","doi-asserted-by":"publisher","DOI":"10.1038\/s41586-024-07198-2"},{"key":"2026041100133766000_2025.02.18.638828v2.14","doi-asserted-by":"publisher","DOI":"10.1107\/S2059798321012122"},{"key":"2026041100133766000_2025.02.18.638828v2.15","doi-asserted-by":"publisher","DOI":"10.1002\/prot.26214"},{"key":"2026041100133766000_2025.02.18.638828v2.16","doi-asserted-by":"publisher","DOI":"10.1107\/S2059798319011471"},{"key":"2026041100133766000_2025.02.18.638828v2.17","doi-asserted-by":"publisher","DOI":"10.1107\/S0907444996012255"},{"key":"2026041100133766000_2025.02.18.638828v2.18","doi-asserted-by":"publisher","DOI":"10.1107\/S2059798323002413"},{"key":"2026041100133766000_2025.02.18.638828v2.19","doi-asserted-by":"publisher","DOI":"10.1038\/nprot.2008.91"},{"key":"2026041100133766000_2025.02.18.638828v2.20","doi-asserted-by":"publisher","DOI":"10.1107\/S0907444903009922"},{"key":"2026041100133766000_2025.02.18.638828v2.21","doi-asserted-by":"publisher","DOI":"10.1107\/S090744490705024X"},{"key":"2026041100133766000_2025.02.18.638828v2.22","doi-asserted-by":"publisher","DOI":"10.1093\/nar\/gkad1019"},{"key":"2026041100133766000_2025.02.18.638828v2.23","doi-asserted-by":"publisher","DOI":"10.1107\/S0907444911047834"},{"key":"2026041100133766000_2025.02.18.638828v2.24","doi-asserted-by":"publisher","DOI":"10.1073\/pnas.96.10.5482"},{"key":"2026041100133766000_2025.02.18.638828v2.25","doi-asserted-by":"publisher","DOI":"10.1002\/prot.22540"},{"key":"2026041100133766000_2025.02.18.638828v2.26","doi-asserted-by":"publisher","DOI":"10.1038\/s41587-023-01704-z"},{"key":"2026041100133766000_2025.02.18.638828v2.27","doi-asserted-by":"crossref","unstructured":"Zhang, Y. et al. Distance-AF: Modifying Predicted Protein Structure Models by Alphafold2 with User-Specified Distance Constraints (2023).","DOI":"10.1101\/2023.12.01.569498"},{"key":"2026041100133766000_2025.02.18.638828v2.28","doi-asserted-by":"publisher","DOI":"10.1038\/s41586-024-07215-4"},{"key":"2026041100133766000_2025.02.18.638828v2.29","doi-asserted-by":"publisher","DOI":"10.1107\/S205979832300102X"},{"key":"2026041100133766000_2025.02.18.638828v2.30","doi-asserted-by":"publisher","DOI":"10.1038\/s42256-024-00873-z"},{"key":"2026041100133766000_2025.02.18.638828v2.31","year":"2024","journal-title":"Solving Inverse Problems in Protein Space Using Diffusion-Based Priors"},{"key":"2026041100133766000_2025.02.18.638828v2.32","year":"2025","journal-title":"Inverse problems with experiment-guided AlphaFold"},{"key":"2026041100133766000_2025.02.18.638828v2.33","doi-asserted-by":"publisher","DOI":"10.1038\/s41592-024-02272-z"},{"key":"2026041100133766000_2025.02.18.638828v2.34","doi-asserted-by":"publisher","DOI":"10.1101\/2025.01.12.632630"},{"key":"2026041100133766000_2025.02.18.638828v2.35","doi-asserted-by":"publisher","DOI":"10.1038\/s41586-023-06832-9"},{"key":"2026041100133766000_2025.02.18.638828v2.36","doi-asserted-by":"publisher","DOI":"10.1371\/journal.pcbi.1012253"},{"key":"2026041100133766000_2025.02.18.638828v2.37","doi-asserted-by":"publisher","DOI":"10.1103\/PRXLife.2.043001"},{"key":"2026041100133766000_2025.02.18.638828v2.38","doi-asserted-by":"publisher","DOI":"10.1038\/s41467-024-46715-9"},{"key":"2026041100133766000_2025.02.18.638828v2.39","doi-asserted-by":"publisher","DOI":"10.1371\/journal.pcbi.1010483"},{"key":"2026041100133766000_2025.02.18.638828v2.40","first-page":"e65211","article-title":"High-Throughput Screening to Obtain Crystal Hits for Protein Crystallography","volume":"193","year":"2023","journal-title":"Journal of Visualized Experiments: JoVE"},{"key":"2026041100133766000_2025.02.18.638828v2.41","doi-asserted-by":"publisher","DOI":"10.1126\/science.abo7201"},{"key":"2026041100133766000_2025.02.18.638828v2.42","doi-asserted-by":"publisher","DOI":"10.1126\/science.abf7945"},{"key":"2026041100133766000_2025.02.18.638828v2.43","doi-asserted-by":"publisher","DOI":"10.1002\/appl.202400192"},{"key":"2026041100133766000_2025.02.18.638828v2.44","doi-asserted-by":"crossref","unstructured":"Dalton, K. M. , Greisman, J. & Hekstra, D. A unifying bayesian framework for merging x-ray diffraction data. Nature Communications 7764 (2022).","DOI":"10.1101\/2021.01.05.425510"},{"key":"2026041100133766000_2025.02.18.638828v2.45","doi-asserted-by":"publisher","DOI":"10.1038\/s41592-020-01049-4"},{"key":"2026041100133766000_2025.02.18.638828v2.46","doi-asserted-by":"publisher","DOI":"10.1038\/s41592-024-02304-8"},{"key":"2026041100133766000_2025.02.18.638828v2.47","doi-asserted-by":"publisher","DOI":"10.1038\/s41592-023-02033-4"},{"key":"2026041100133766000_2025.02.18.638828v2.48","doi-asserted-by":"publisher","DOI":"10.1107\/S2059798318002425"},{"key":"2026041100133766000_2025.02.18.638828v2.49","doi-asserted-by":"publisher","DOI":"10.1101\/2024.01.20.576370"},{"key":"2026041100133766000_2025.02.18.638828v2.50","doi-asserted-by":"publisher","DOI":"10.1038\/s41467-022-31748-9"},{"key":"2026041100133766000_2025.02.18.638828v2.51","doi-asserted-by":"publisher","DOI":"10.1107\/S0021889807021206"},{"key":"2026041100133766000_2025.02.18.638828v2.52","doi-asserted-by":"publisher","DOI":"10.1107\/S2059798323001596"},{"key":"2026041100133766000_2025.02.18.638828v2.53","doi-asserted-by":"publisher","DOI":"10.1107\/S2059798324011999"},{"key":"2026041100133766000_2025.02.18.638828v2.54","doi-asserted-by":"publisher","DOI":"10.1002\/pro.3353"},{"key":"2026041100133766000_2025.02.18.638828v2.55","doi-asserted-by":"publisher","DOI":"10.1038\/355472a0"},{"key":"2026041100133766000_2025.02.18.638828v2.56","unstructured":"Jendrusch, M. , Korbel, J. O. & Sadiq, S. K. AlphaDesign: A de novo protein design framework based on AlphaFold. BioRxiv 10.1101\/2021.10.11.463937 (2021)."},{"key":"2026041100133766000_2025.02.18.638828v2.57","doi-asserted-by":"publisher","DOI":"10.1126\/science.add1964"},{"key":"2026041100133766000_2025.02.18.638828v2.58","doi-asserted-by":"publisher","DOI":"10.1126\/science.adq1741"},{"key":"2026041100133766000_2025.02.18.638828v2.59","doi-asserted-by":"publisher","DOI":"10.1101\/2021.10.04.463034"},{"key":"2026041100133766000_2025.02.18.638828v2.60","doi-asserted-by":"publisher","DOI":"10.1517\/17460441.2014.872623"},{"key":"2026041100133766000_2025.02.18.638828v2.61","doi-asserted-by":"publisher","DOI":"10.1107\/S2059798320014746"},{"key":"2026041100133766000_2025.02.18.638828v2.62","doi-asserted-by":"publisher","DOI":"10.1107\/S2059798323001602"},{"key":"2026041100133766000_2025.02.18.638828v2.63","doi-asserted-by":"publisher","DOI":"10.1107\/S2059798315013236"},{"key":"2026041100133766000_2025.02.18.638828v2.64","doi-asserted-by":"publisher","DOI":"10.1107\/S0108767390005529"},{"key":"2026041100133766000_2025.02.18.638828v2.65","unstructured":"Fletcher, R. Practical Methods of Optimization 2nd edn (1987)."},{"key":"2026041100133766000_2025.02.18.638828v2.66","first-page":"1","article-title":"A purely algebraic justification of the Kabsch-Umeyama algorithm","volume":"124","year":"2019","journal-title":"Journal of research of the National Institute of Standards and Technology"},{"key":"2026041100133766000_2025.02.18.638828v2.67","unstructured":"Schr\u00f6dinger, LLC. The PyMOL molecular graphics system, version 1.8 (2015)."},{"key":"2026041100133766000_2025.02.18.638828v2.68","unstructured":"Moorhen: A web browser molecular graphics program based on the coot desktop program (2025). URL https:\/\/github.com\/moorhen-coot\/Moorhen."},{"key":"2026041100133766000_2025.02.18.638828v2.69","doi-asserted-by":"publisher","DOI":"10.1002\/pro.3330"}],"container-title":[],"original-title":[],"link":[{"URL":"https:\/\/syndication.highwire.org\/content\/doi\/10.1101\/2025.02.18.638828","content-type":"unspecified","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2026,6,24]],"date-time":"2026-06-24T05:11:19Z","timestamp":1782277879000},"score":1,"resource":{"primary":{"URL":"http:\/\/biorxiv.org\/lookup\/doi\/10.1101\/2025.02.18.638828"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2025,2,21]]},"references-count":69,"URL":"https:\/\/doi.org\/10.1101\/2025.02.18.638828","relation":{"is-preprint-of":[{"id-type":"doi","id":"10.1038\/s41592-026-03047-4","asserted-by":"subject"}]},"subject":[],"published":{"date-parts":[[2025,2,21]]},"subtype":"preprint"}}