The Pulse
Biohub Expands Its Virtual Cell Effort to $1.8 Billion
Biohub and government, technology and scientific partners announced a $1.8 billion effort to generate biological data for AI models that predict how cells respond to changes. Commercial partners will have a one-year exclusive window on data
AI.info Team · Written with AI ·
Biohub’s $1.8 billion push to model cell behavior aims to create an open scientific resource, while commercial partners get a one-year head start on the data they develop. Announced on October 7, 2026, the expanded effort brings together Biohub, the U.S. Department of Energy, the National Institutes of Health, Google DeepMind, Isomorphic Labs, Meta and scientific organizations. Its goal is to build data and tools that let AI models predict how cells respond to changes, so researchers can test some ideas digitally before committing to costly laboratory experiments.
A $1.8 billion effort built from different contributions
The total combines funding, data, computing and measurement technology; it is not a single cash pot. The Department of Energy plans to contribute more than $500 million over five years through its Genesis Mission, supporting biological measurement, modeling and computing. Google DeepMind, Isomorphic Labs and Meta are committing a combined $300 million to the Virtual Biology Initiative.
Biohub’s existing $500 million commitment is divided between $400 million for new measurement technologies and $100 million for research outside Biohub. The $100 million is part of that $500 million commitment, not an additional contribution.
NIH’s contribution is different: it will coordinate relevant datasets, repositories and knowledge bases developed through more than $500 million in earlier federal investment. Biohub will work with the agency to standardize those resources for AI training.
The missing ingredient is measured biology
The project is not simply a bid to train a larger model. Its first phase aims to gather different kinds of information about cells and their responses to interventions, extending measurements across more cell types and conditions than researchers have studied so far. The federal effort will draw on national-laboratory capabilities including exascale computing, X-ray and neutron scattering, cryo-electron microscopy and tomography, and autonomous laboratories.
“An accurate predictive model of biology could dramatically accelerate scientific discovery by enabling scientists to perform experiments digitally.”
— Biohub Head of Science Alex Rives
Much of the evidence needed to model living cells must first be produced through experiments and measurement. Biohub and its partners are seeking to create and standardize that evidence, not announcing a working whole-cell simulator.
Companies get a year before their data opens
Commercial partners will have one year of exclusive access to the data they develop before it is shared publicly. That arrangement applies to partner-developed data; it does not mean all data generated through the initiative will necessarily be subject to the same period of exclusive access. Rives said the temporary advantage is intended to give companies a reason to contribute while keeping the resulting data on a path to broader scientific use.
The access policy makes the initiative’s open-resource goal dependent in part on which data partners develop and when those data are released. The announcement describes the eventual resource as open, while the reported one-year window gives participating companies early access to their own contributions.
The first test comes after the first large dataset
The immediate scientific question is whether more biological data will produce measurable gains in model predictions. Rives told Axios that within a year of the first large-scale dataset, researchers should be able to train models, assess their capabilities and identify which additional kinds of data improve them. That evaluation will offer an early check on the central bet behind the $1.8 billion coalition: that enough carefully measured evidence can make cell behavior predictable enough to guide experiments.