CoCoFold2: scalable latent refinement of diffusion-based protein structure predictions from limited-particle cryo-EM data
Abstract
Motivation: Limited-particle cryo-EM data may be insufficient for reliable de novo model building yet still provide structural constraints for correcting a corresponding sequence-based prediction. Extending this refinement setting to long protein sequences while still achieving meaningful improvements over the initial prediction remains challenging.
Results: We present CoCoFold2, a deterministic latent-refinement framework that learns a target-specific latent perturbation through a differentiable particle forward model while keeping the pretrained Protenix-v1 diffusion prior and a single diffusion realization fixed. To support long-sequence refinement, CoCoFold2 implements component-parallel refinement, as demonstrated on the four-chain, 1,683-residue merozoite surface protein 1 (MSP-1) assembly (PDB ID 6ZBH) and the 4,392-residue calcium-bound transient receptor potential melastatin 8 (TRPM8) tetramer (PDB ID 6O77). We evaluated the fixed-stochasticity latent-refinement procedure on an all-eligible post-cutoff benchmark comprising nine public protein-only EMPIAR datasets and one in-house dataset. Relative to unrefined Protenix-v1 predictions, fixed-stochasticity latent refinement yielded lower backbone root-mean-square deviation (RMSD) and higher completeness. It remained competitive with CryoAtom2 and CoCoFold, which served as map-based and prior-guided comparison methods, respectively. In three selected targets, the fixed-stochasticity procedure outperformed controls that resampled diffusion stochasticity or fine-tuned diffusion-module parameters.
Availability and Implementation: Source code is available at https://github.com/jwliaomath/CoCoFold2.
Contact: humingxu@smart.org.cn; clbao@mail.tsinghua.edu.cn.
References
Abramson, J., Adler, J., Dunger, J., Evans, R., Green, T., Pritzel, A., Ronneberger, O., Willmore, L., Ballard, A.J., Bambrick, J., Bodenstein, S.W., Evans, D.A., Hung, C.-C., O'Neill, M., Reiman, D., Tunyasuvunakool, K., Wu, Z., Zemgulytė, A., Arvaniti, E., Beattie, C., Bertalli, O., Bridgland, A., Cherepanov, A., Congreve, M., Cowen-Rivers, A.I., Cowie, A., Figurnov, F.B., Fuchs, H., Gladman, R., Jain, Y.A., Khan, K.M.R. Low, Perlin, K., Potapenko, A., Savy, P., Singh, S., Stecula, A., Thillaisundaram, C., Tong, S., Yakneen, S., Zhong, E.D., Zielinski, M., Zidek, A., Bapst, V., Kohli, P., Jaderberg, M., Hassabis, D., and Jumper, J.M. Accurate structure prediction of biomolecular interactions with AlphaFold 3. Nature, 630(8016):493–500, 2024. doi:10.1038/s41586-024-07487-w
Cossio, P. and Hummer, G. Bayesian analysis of individual electron microscopy images: towards structures of dynamic and heterogeneous biomolecular assemblies. Journal of Structural Biology, 184(3):427–437, 2013. doi:10.1016/j.jsb.2013.10.006
Fadini, A., Li, M., McCoy, A.J., Banjara, S., Okumura, H., Napier, E., Fontana, P., Khan, A.R., Jovine, L., Terwilliger, T.C., Read, R.J., Hekstra, D.R., and AlQuraishi, M. AlphaFold as a Prior: Experimental Structure Determination Conditioned on a Pretrained Neural Network. Nature Methods, 23(4):785–795, 2026. doi:10.1038/s41592-026-03047-4
Jamali, K., Kall, L., Zhang, R., Brown, A., Kimanius, D., and Scheres, S.H.W. Automated model building and protein identification in cryo-EM maps. Nature, 628(8007):450–457, 2024. doi:10.1038/s41586-024-07215-4
Jumper, J., Evans, R., Pritzel, A., Green, T., Figurnov, M., Ronneberger, O., Tunyasuvunakool, R., Bates, R., Žídek, A., Potapenko, A., Bridgland, A., Meyer, C., Kohl, S.A.A., Ballard, A.J., Cowie, A., Romera-Paredes, B., Nikolov, S., Jain, R., Adler, J., Back, T., Petersen, S., Reiman, D., Clancy, E., Zielinski, M., Steinegger, M., Pacholska, M., Berghammer, T., Bodenstein, S., Silver, D., Vinyals, O., Senior, A.W., Kavukcuoglu, K., Kohli, P., and Hassabis, D. Highly accurate protein structure prediction with AlphaFold. Nature, 596(7873):583–589, 2021. doi:10.1038/s41586-021-03819-2
Li, M., Han, J., Cossio, P., and Wu, L. Robust inference-time steering of protein diffusion models via embedding optimization. arXiv, 2026. https://arxiv.org/abs/2602.05285
Liao, J., Zheng, D., Zhang, H., Zhang, L., Hu, M., and Bao, C. Fine-tuning AlphaFold with limited cryo-EM observations. Communications Chemistry, 9(1):95, 2026. doi:10.1038/s42004-026-01899-7
Lyumkis, D. Challenges and opportunities in cryo-EM single-particle analysis. Journal of Biological Chemistry, 294(13):5181–5197, 2019. doi:10.1074/jbc.REV118.005602
Maddipatla, A., Rzayev, A., Pegoraro, M., Pacesa, M., Schanda, P., Marx, A., Vedula, S., and Bronstein, A.M. Inference-time optimization for experiment-grounded protein ensemble generation. arXiv, 2026a. https://arxiv.org/abs/2602.24007
Maddipatla, A., Sellam Bojan, N., Bojan, M., Masalitin, V., Vedula, S., Schanda, P., Marx, A., and Bronstein, A.M. Experiment-guided AlphaFold3 resolves measurement-consistent protein ensembles. Nature Biotechnology, 2026b. doi:10.1038/s41587-026-03166-5
Naydenova, K. and Russo, C.J. Measuring the effects of particle orientation to improve the efficiency of electron cryomicroscopy. Nature Communications, 8(1):629, 2017. doi:10.1038/s41467-017-00782-3
Raghu, R., Levy, A., Wetzstein, G., and Zhong, E.D. Multiscale guidance of protein structure prediction with heterogeneous cryo-EM data. Advances in Neural Information Processing Systems, 38, 2025. doi:10.52202/085713-0344
Shenoy, J., Astore, M., Levy, A., Poitevin, F., Hanson, S.M., and Wetzstein, G. Modeling atomic conformational ensembles of proteins via test-time supervision of Boltz-2 on cryo-EM density maps. arXiv, 2026. https://arxiv.org/abs/2605.09832
Stagg, S.M., Noble, A.J., Spilman, M., and Chapman, M.S. ResLog plots as an empirical metric of the quality of cryo-EM reconstructions. Journal of Structural Biology, 185(3):418–426, 2014. doi:10.1016/j.jsb.2013.12.010
Su, B., Zhu, L., Li, W., Wang, W., Liang, J., Zhao, K., Peng, Z., and Yang, J. CryoAtom2: inverse folding-inspired deep learning enables accurate model building of protein-nucleic acid complexes in cryo-EM. 2026. https://langtaosha.org.cn/lts/en/preprint/view/185
Tang, W.S., Silva-Sanchez, D., Giraldo-Barreto, J., Carpenter, B., Hanson, S.M., Barnett, A.H., Thiede, E.H., and Cossio, P. Ensemble reweighting using cryo-EM particle images. The Journal of Physical Chemistry B, 127(24):5410–5421, 2023. doi:10.1021/acs.jpcb.3c01087
Terwilliger, T.C., Poon, B.K., Afonine, P.V., Schlicksup, C.J., Croll, T.I., Millán, C., Richardson, J.S., Read, R.J., and Adams, P.D. Improved AlphaFold modeling with implicit experimental information. Nature Methods, 19(11):1376–1382, 2022. doi:10.1038/s41592-022-01645-6
Terwilliger, T.C., Liebschner, D., Croll, T.I., Williams, C.J., McCoy, A.J., Poon, B.K., Afonine, P.V., Oeffner, R.D., Richardson, J.S., Read, R.J., and Adams, P.D. AlphaFold predictions are valuable hypotheses and accelerate but do not replace experimental structure determination. Nature Methods, 21(1):110–116, 2024. doi:10.1038/s41592-023-02087-4
Wang, X., Zhu, H., Terashi, G., Taluja, M., and Kihara, D. DiffModeler: large macromolecular structure modeling for cryo-EM maps using a diffusion model. Nature Methods, 21:2307–2317, 2024. doi:10.1038/s41592-024-02479-0
Zhang, Y., Gong, C., Zhang, H., Ma, W., Liu, Z., Chen, X., Guan, J., Wang, L., Yang, Y., Xia, Y., and Xiao, W. Protenix-v1: Toward high-accuracy open-source biomolecular structure prediction. bioRxiv, 2026. doi:10.64898/2026.02.05.703733
Zhong, E.D., Bepler, T., Berger, B., and Davis, J.H. CryoDRGN: reconstruction of heterogeneous cryo-EM structures using neural networks. Nature Methods, 18(2):176–185, 2021. doi:10.1038/s41592-020-01049-4
Metrics
DOI:
Submission ID:
Downloads
Additional Files
Supplemental File(s)
Posted
How to Cite
Download Citation
Declaration of Competing Interests
The authors declare no competing interests to disclose.
Copyright
The copyright holder for this preprint is the author/funder.
This work is licensed under a Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License.