Reproduce “RNADyn: A Benchmark for Generating and Understanding RNA Dynamics” on OpenResearch

Turn your favorite coding agents into research agents that review literature, develop hypotheses, and run experiments.

Universal app for Apple silicon and Intel. Requires macOS 11 or later.

About this paper

Yiming Huang, Lennart Bastian, Hanqun Cao, Luis Vollmers, Tolga Birdal

The authors introduce RNADynNet, which generates all-atom RNA trajectories from a starting conformer and predicts residue-level dynamics from one conformer. Its shared backbone uses trajectory alignment and physical grounding, trained on RNADynBench simulations. On test_flex, grounding lowers maximum linkage deviation from 9.1171 to 4.2400 angstroms and improves RMSF correlation; single-conformer prediction exceeds BackFlip, within the 100-ns, primarily isolated-RNA scope.

Summary by alphaXiv.

Tip: change arxiv.org to autoarxiv.org in any arXiv abstract link to open the paper here.