Biohub leads $1.8 billion virtual-cell AI initiative backed by Google DeepMind, Meta, and Isomorphic Labs

Zuckerberg- and Chan-founded Biohub is coordinating a $1.8 billion effort to build AI models and datasets for a 'virtual cell', with $300 million from DeepMind, Meta, and Isomorphic Labs plus over $500 million each from the DOE and NIH.

Biohub, the biomedical nonprofit founded by Mark Zuckerberg and Priscilla Chan, is coordinating a $1.8 billion effort to build AI datasets and models for a 'virtual cell' that lets researchers simulate biological questions digitally and predict cell behavior, potentially speeding drug development. Google DeepMind, Meta, and Isomorphic Labs are jointly contributing $300 million. The US Department of Energy plans to invest more than $500 million over five years in lab measurements and compute, while the National Institutes of Health is contributing datasets and repositories backed by more than $500 million in prior federal funding, which Biohub will standardize for training. Commercial funders receive one year of exclusive data access before release, and a first dataset is expected in about a year. The two reports agree on all figures; they differ only in emphasis, with The Verge framing the effort around AI datasets for a virtual cell and The Decoder framing it around AI models that predict cell behavior.

• Biohub, the nonprofit founded by Mark Zuckerberg and Priscilla Chan, is coordinating a $1.8 billion biomedical AI initiative.
• Google DeepMind, Meta, and Isomorphic Labs are jointly contributing $300 million to Biohub.
• The goal is AI datasets and models for a 'virtual cell' that simulates biological questions digitally and predicts cell behavior, potentially accelerating drug development.
• The US Department of Energy plans to invest more than $500 million over five years in lab measurements and compute.
• The National Institutes of Health is contributing datasets and repositories backed by more than $500 million in prior federal funding, which Biohub will standardize for training.
• Commercial funders get one year of exclusive access to the data before it is released.
• A first dataset is expected in about a year.