Reproduce “BridgeDock: Taming Generative Co-Folding Prior for Molecular Docking with Diffusion Bridge” 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

Shikun Feng, Zipeng Yang, Qiuyi Li, Mingze Yin, Xiaoyuan Zhang, Tingjun Hou, Yiheng Zhu

BridgeDock predicts flexible molecular docking by guiding generation from an unbound receptor–ligand arrangement toward a bound complex while adapting a pretrained AF3-style denoiser. On PoseBusters, one-step BridgeDock reaches 90.77% lDDT, versus 84.79% for 20-step Protenix; ligand RMSD below 2 Å is 69.86% versus 68.61%. These benchmark results show comparable or better quality with fewer denoising steps, not measured wall-clock speedup.

Summary by alphaXiv.

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