Three of Seattle’s top research institutions — the Allen Institute, the University of Washington, and Fred Hutch Cancer Center — are launching a roughly $95 million initiative, AI BioDesign, to use artificial intelligence to design proteins and genes that don’t exist in nature.
The five-year project is funded by the Fund for Science and Technology, created by the estate of Microsoft co-founder Paul Allen. It will generate large purpose-built datasets and train AI models to invent novel biological parts — with results shared freely to help researchers develop new medicines and materials.
A different bet
AI BioDesign is deliberately wagering on narrow, task-specific models — trained on data generated for a particular design problem — rather than the giant general-purpose models much of the field is chasing. Leading it is David Baker, the Nobel-winning protein-design pioneer and director of UW’s Institute for Protein Design. “For the first time, the speed of AI is beginning to match the experimental power of synthetic biology,” he said.
Why it matters
Designing custom proteins and genes from scratch could accelerate everything from new drugs and vaccines to enzymes and materials. By pairing AI-driven design with the lab “muscle” to make and test the results — and releasing the models openly — the effort aims to speed discovery across many labs, not just its own. It’s a large, coordinated bet that purpose-built AI, closely tied to experiments, is the faster route to real-world biological breakthroughs.