RNADyn: A Benchmark for Generating and Understanding RNA Dynamics

Yiming Huang, Lennart Bastian, Hanqun Cao, Luis Vollmers, Tolga Birdal|October 2, 2026arXiv

Key Takeaway

A unified deep learning approach can both generate realistic RNA dynamics trajectories and predict dynamics fingerprints from static structures, bridging two previously separate tasks and improving physical accuracy through explicit physical constraints.

Summary

This paper introduces RNADynBench, a large-scale benchmark of 2,585 RNA molecular dynamics simulations, and RNADynNet, a unified model that generates realistic RNA trajectories and extracts dynamics information from single structures.

data architecture evaluation

Key Terms

molecular-dynamics rmsf conformational-changes coordinate-denoising physics-informed-neural-network