Microsoft Research introduced Quine, an AI system for biology. It combines a multimodal world model with a harness for orchestration and reasoning. The model spans genomics, proteins, chemistry, cell state and bioimaging. It connects models, tools, literature and researchers in one loop. A collaboration with Broad Institute used Quine to prioritize cancer compounds. Top-ranked candidates shifted tumor cell states in wet-lab assays. The process took one weekend instead of months. Some compounds showed unexpected mechanisms of action. A Quine Fellows program will give scientists access to the system. Quine is experimental and limited to research use. Outputs may be incomplete or inaccurate and need validation. Microsoft expects to expand access via products like Microsoft Discovery.
What changed
Biology research relied on slow, sequential wet-lab experiments without computational prioritization.
What it unlocks
Prioritizing compounds for wet-lab testing using a multimodal AI world model of biology.
- one weekend to prioritize compounds
- thousands of compounds predicted
What you need to act on it
- research use only
- not for clinical or medical use
- requires qualified researcher review
Sources