Biohub's Virtual Biology Initiative: About $1.8B in Funding and Resources

An April initiative gets a government and industry partner list

Biohub, a biomedical research nonprofit, launched the Virtual Biology Initiative on April 29, 2026 as a five-year effort to produce the datasets and measurement tools needed to train predictive models of cells. It committed $500 million of its own: about $100 million to fund data generation by outside groups and about $400 million for its own data and new instruments. [3]

On October 7, Biohub announced from Redwood City, California that the initiative had grown to about $1.8 billion "in funding, data, computation, and new measurement technology." Reuters reported the same figure that morning, Eastern time. Biohub's head of science, Alex Rives, called building a virtual cell "one of the most important challenges for the next era of science" and invited outside scientists to join. [1][2]

Who puts in what

The Department of Energy committed more than $500 million over five years through its Genesis Mission, including time on national laboratory supercomputers, X-ray and neutron scattering, cryo-electron microscopy and autonomous labs. The National Institutes of Health agreed to coordinate datasets and repositories built with more than $500 million of earlier federal funding, which Biohub will standardize for AI training. Google DeepMind, Isomorphic Labs and Meta together committed $300 million; the release does not split it among them. [1][2]

The release also named the Allen Institute, the Broad Institute, Gladstone Institutes, the Human Cell Atlas, the Human Protein Atlas and the Wellcome Sanger Institute as data partners, NVIDIA as a provider of computing and software, and Renaissance Philanthropy as a funder of further data generation. [1]

The headline figure is not $1.8 billion of new cash. Biohub's $500 million is the April pledge, and the NIH contribution is coordination of data that already exists. The new commitments in October are the Energy Department's money and facilities and the $300 million from the three companies, which together come to a little over $800 million. [1][3]

Cell responses, open after an embargo

The partners plan to measure how cells respond to interventions across more cell types and conditions than have been studied, and to build shared standards, common identifiers and a single point of access across their datasets. Rives told Reuters that today's cell datasets hold hundreds of millions of cells while accurate models will need billions and eventually trillions, that the first dataset was expected in about a year, and that he expected accurate predictive models within five years. [1][2]

Biohub describes the result as an open resource for the research community but publishes no license or access terms. Rives told Reuters that commercial funders get embargo periods to work on the data before it is released, and that government-funded work would carry no such restriction. [1][2]

The initiative is a bet on measurement before models. A month earlier, Google DeepMind's AlphaGenome Atlas went the other way, publishing model predictions for every possible single-letter change in the human genome. The Virtual Biology Initiative, which Google DeepMind also funds, is meant to produce the experimental cell data that models of this kind are trained on and checked against. [1][2]