OpenComplex

Open-source platform for end-to-end prediction of biomolecular 3D structures including protein, RNA, and protein-RNA complexes. Built on architectures of AlphaFold2 and AlphaFold-Multimer. Part of BAAI's WuJie series focused on life science AI. Provides a trainable PyTorch framework for developing protein, RNA, and complex models.

Generative foundation model for biomolecular conformational ensemble sampling. Integrates a unified graph representation with diffusion-based sampling through the FloydNetwork architecture, operating across atomic, residue, and motif hierarchies. Samples conformational landscapes in hours instead of the weeks required by traditional molecular dynamics. Accurately reproduces protein and RNA ensembles, captures functionally relevant transitions (allostery, induced-fit, cryptic pockets), and scales to assemblies exceeding 15,000 residues.