Advanced molecular structure prediction for drug discovery
Chai Discovery, founded in 2024, introduces Chai-1, a multi-modal foundation model designed for molecular structure prediction. This model excels in various tasks pertinent to drug discovery, achieving state-of-the-art performance. Chai-1 enables unified predictions for proteins, small molecules, DNA, RNA, and covalent modifications. It can operate in single sequence mode without requiring multiple sequence alignments (MSAs), maintaining high accuracy. Notably, it predicts multimer structures using only single sequences, outperforming MSA-based models like AlphaFold-Multimer. The model is accessible for free through a web interface, supporting commercial applications in drug discovery, and its code is available under the Apache 2.0 License. The Chai Discovery team comprises experts from leading AI research organizations, emphasizing collaboration with the research and industrial communities to enhance the ecosystem.