Will be released soon. Please stay tuned.

OpenComplex2

OpenComplex2 (OC2) is a generative foundation model that bridges static structure prediction and dynamic ensemble modeling of biomolecular systems. Unlike traditional structure prediction methods that focus solely on a single conformation, OC2 can efficiently sample thermodynamically relevant conformational ensembles, providing insights into molecular function that static structures alone cannot capture.

Key Features

📊 Equilibrium Ensemble Sampling

• NMR-Validated Distributions: Accurately reproduces experimentally determined conformational ensembles
• MD-Comparable Sampling: Achieves sampling quality comparable to millisecond-scale molecular dynamics simulations at a fraction of the computational cost

🧬 High-Accuracy Structure Prediction

• State-of-the-Art Accuracy: Competitive with specialized structure prediction tools on standard benchmarks Ultra-Large Assemblies: Efficiently models assemblies exceeding 15,000 residues through symmetry-aware optimizations
• Precise Small Molecule Modeling: Accurate stereochemical modeling of ligands and identification of multiple binding modes

Citation

If you find our open-sourced code & models helpful to your research, please also consider star🌟 and cite📑 this repo. Thank you for your support!

@article{opencomplex2025towards,
  title={Towards Unraveling Biomolecular Conformational Landscapes with a Generative Foundation Model},
  author={OpenComplex Team},
  journal={bioRxiv},
  pages={2025--05},
  year={2025},
  publisher={Cold Spring Harbor Laboratory}
}

Contact Information

For help or issues using the repos, please submit a GitHub issue.

For other communications, please contact Qiwei Ye (qwye@baai.ac.cn).