
Google DeepMind's new AI system can design novel proteins to bind with high affinity to target molecules, potentially revolutionizing drug discovery.
Following the success of AlphaFold, Google DeepMind has introduced AlphaProteo, an AI system capable of designing entirely new proteins that bind to specific target molecules. This is a critical step in developing new treatments for diseases like cancer and viral infections where existing binders are ineffective.
AlphaProteo was trained on vast amounts of protein structure data from the Protein Data Bank. It can generate candidate binders that are far more potent than those created using traditional methods. In lab tests, the system designed binders for the SARS-CoV-2 spike protein that were significantly more effective than existing antibodies.
The implications for the pharmaceutical industry are massive. By automating the design of high-affinity binders, researchers can cut years off the early-stage drug discovery process. DeepMind plans to share these tools with the scientific community to accelerate biological research across the board.

