RNADyn: A Benchmark for Generating and Understanding RNA Dynamics

2026-10-02

AI Summary

This study addresses the scarcity of RNA dynamic data and the disconnect between trajectory generation and kinetic analysis by proposing a unified benchmark and modeling framework. Methodologically, it constructs a shared backbone network that integrates coordinate denoising, single-frame-to-trajectory alignment, and physics-grounded mechanisms. Combined with all-atom molecular dynamics (MD) simulations, this approach enables trajectory generation from single conformations while simultaneously extracting kinetic features. Experimental results demonstrate that the generated trajectories achieve RMSF correlations of 0.875 and 0.766, respectively. Furthermore, the prediction accuracy from single conformations is comparable to MD simulation outcomes. By effectively bridging the gap between generative modeling and representation learning, this work provides a robust foundation for advancing RNA conformational dynamics research.

Abstract

Ribonucleic acid (RNA) functions through conformational changes that are not fully captured by static structures. However, large-scale standardized RNA dynamics data remain limited, and existing approaches typically treat trajectory generation and dynamics understanding as separate objectives. Here, we introduce RNADynBench, a standardized RNA molecular dynamics (MD) benchmark with 2585 quality-controlled 100-ns all-atom trajectories and leakage-controlled splits. Building on RNADynBench, we develop RNADynNet, a unified model for RNA dynamics learning that uses a shared backbone for both trajectory generation and dynamics fingerprint extraction from a single conformer. It combines coordinate denoising, single-frame-to-trajectory alignment, and physical grounding to connect all-atom trajectory generation with dynamics representation learning. Physical grounding improves both generated dynamics and the physical information recoverable from these fingerprints. Across both test sets, including the high-flexibility challenge set, the generated trajectories achieve RMSF correlations of 0.875 and 0.766, while single-conformer predictions show comparable agreement with MD-derived dynamics. RNADynBench and RNADynNet together establish a benchmark and unified modeling framework for generating and understanding RNA dynamics.

Problem

• RNA dynamics
• molecular dynamics benchmark
• trajectory generation
• dynamics understanding

Innovation

• RNA dynamics
• molecular dynamics benchmark
• unified model
• trajectory generation
• physical grounding