Google DeepMind's Co-Scientist AI Accelerates Liver Fibrosis Drug Discovery
A new multi-agent AI system called Co-Scientist helps researchers uncover overlooked drug-repurposing candidates for treating chronic liver disease. One identified treatment successfully blocks 91% of a scarring-linked response in laboratory tests.
Google DeepMind introduces Co-Scientist, a multi-agent AI system designed to act as a collaborative partner for scientific researchers. The tool aims to accelerate the discovery process by analyzing vast amounts of existing literature and identifying connections that human scientists might overlook.
In a recent case study, Stanford Professor Gary Peltz uses Co-Scientist to search for new treatments for liver fibrosis. The AI system highlights overlooked drug-repurposing candidates, identifying a specific treatment that successfully blocks 91% of a scarring-linked response during laboratory tests.
These promising results appear in the journal Advanced Science and point toward innovative gene-regulating approaches for chronic liver disease. Peltz describes the AI as a collaborator that reads all available biomedical science and possesses the reasoning capabilities to find the critical connections researchers are currently missing.