DeepMind AlphaFold 3 Predicts DNA, RNA, and Chemical Compound Interactions

Google DeepMind has unveiled AlphaFold 3, the latest version of its AI system that has already revolutionized biology. While AlphaFold 2 focused entirely on predicting protein structures, AlphaFold 3 expands its capabilities to model interactions between all of life's molecules.

The model can simulate how proteins interact with DNA, RNA, chemical ligands, and ions, representing a monumental leap for drug discovery.

Modeling the Entire Molecular Ecosystem AlphaFold 3 utilizes a diffusion-based architecture similar to AI image generators to predict the 3D structures of molecular complexes. - **DNA and RNA:** Researchers can now see exactly how proteins bind to genetic sequences, opening up new paths for gene therapies. - **Drug Interactions:** Chemists can model how potential drug compounds bind to target receptors, accelerating preclinical trials.

"This is the most complete picture of biological machinery we have ever had," said Demis Hassabis, CEO of Google DeepMind. The tool has been made available to researchers worldwide through the AlphaFold Server, ensuring global collaboration.