Running summarizing tools on a new article
This is the first time this article is requested and our AI summarizing tools have never been run on it. We can run our tools now if you click on the button "Run" donw the page but first make sure that it is the right article.
AlphaFold Meets Flow Matching for Generating Protein Ensembles
Bowen Jing, Bonnie Berger, Tommi Jaakkola
Abstract: The biological functions of proteins often depend on dynamic structural ensembles. In this work, we develop a flow-based generative modeling approach for learning and sampling the conformational landscapes of proteins. We repurpose highly accurate single-state predictors such as AlphaFold and ESMFold and fine-tune them under a custom flow matching framework to obtain sequence-conditoned generative models of protein structure called AlphaFlow and ESMFlow. When trained and evaluated on the PDB, our method provides a superior combination of precision and diversity compared to AlphaFold with MSA subsampling. When further trained on ensembles from all-atom MD, our method accurately captures conformational flexibility, positional distributions, and higher-order ensemble observables for unseen proteins. Moreover, our method can diversify a static PDB structure with faster wall-clock convergence to certain equilibrium properties than replicate MD trajectories, demonstrating its potential as a proxy for expensive physics-based simulations. Code is available at https://github.com/bjing2016/alphaflow.
⚠The license of this specific paper does not allow us to build upon its content and the summarizing tools will be run using the paper metadata rather than the full article. However, it still does a good job, and you can also try our tools on papers with more open licenses.