RPDynaFlow is presented, a flow-matching model to generate conformation ensembles of RNA–protein complexes, trained on 600 ns trajectories of molecular dynamics simulation, which could be treated as a rapid and efficient complement to MD trajectories for studying RNA–protein interactions.
Abstract
Conformation ensembles of biomolecules provide the basis for understanding structural transformations and drug design. Deep-learning generative models have advanced protein and small molecule ensemble generation, while RNA–protein complexes remain unaddressed due to the chemical heterogeneity, limited dataset size and the different flexibility scales of RNA and protein components. We present RPDynaFlow, a flow-matching model to generate conformation ensembles of RNA–protein complexes, trained on 600 ns trajectories of molecular dynamics (MD) simulation. The results show our model extends the sampling range of the phase space compared to MD simulation, which could be treated as a rapid and efficient complement to MD trajectories for studying RNA–protein interactions.
Pi-Ensemble (Predicting Interpolated Ensemble), a sequence-guided framework for generating protein conformational ensembles interpolating between two structural anchor states, provides an extensible framework for studying protein flexibility, guiding adaptive sampling, and accelerating mechanistic investigations of pro...
Hassan Nadeem, D. Kleiman, Yu-Ming Zhou et al.· bioRxiv· 0 citations
These approaches improve generalisability, reduce reliance on deep evolutionary information, and enable proteome-scale prediction of RNA-binding residues, providing a route to map and interpret the molecular logic of protein-RNA interactions.
Rozeena Arif, Alfredo Castello· Current Opinion in Structura...· 0 citations
This work assesses ML potentials for exploring RNA conformations using the adenine–adenine dinucleoside monophosphate (ApA) dimer, a fundamental RNA building block, and parametrized ML potentials based on the equivariant MACE architecture and informed by both ab initio and semiempirical property data.
Leonardo Medrano Sandonas, Macarena Tolmos Nehme, L. F. Cofas-Vargas et al.· Journal of Chemical Theory a...· 0 citations
It is shown that intricate RNA structures can be generated with current deep learning tools through accurate de novo design of pseudoknot secondary structures, suggesting that some difficult RNA design tasks may be tractable without first solving RNA 3D structure prediction.
J. Townley, W. Kladwang, David Baker et al.· Science· 1 citation
Recent advances in protein structure prediction, exemplified by AlphaFold, have largely addressed the determination of static structures, one aspect of the protein folding problem. However, predicting folding pathways, by which proteins reach their native states, remains a significant challenge. Here, we present PathFo...
