The $1.8 billion virtual cell: the US government, Biohub and big tech put to the open-data test
On 7 October 2026, Biohub, the US Department of Energy (DOE) and the National Institutes of Health (NIH) announced an expansion of the Virtual Biology Initiative. According to the DOE, its Office of Science, the NIH and Biohub have signed a Memorandum of Understanding to build an open foundation of biological data for predictive models of living systems, the so-called “virtual cell”. Biohub's press release speaks of “nearly $2 billion” in its headline. The body of the release puts the total value at $1.8 billion across funding, data, compute and measurement technologies, and that figure includes resources that already existed. The new money declared is about $500 million from the DOE and $300 million from the companies, plus the $500 million from Biohub that was already announced in April.
The money comes from several directions, and each backer has its own interests. The DOE will contribute more than $500 million over five years through the Genesis Mission, providing autonomous robotic laboratories and exascale supercomputing. Google DeepMind, Isomorphic Labs and Meta are putting in $300 million between them, and they have not said how that amount is split. The NIH is not announcing a new figure. Through the Bio Genesis Mission it coordinates datasets, repositories and knowledge bases built on more than $500 million in federal investment already spent, and it will help Biohub standardise the data for model training.
The promise of open data science still faces some practical questions. The text of the memorandum is not currently public, and the institutional sources have not yet said which licences will govern the datasets or when the data will be released. The Neuron newsletter, citing Axios and Reuters, reports that commercial partners could get one year of exclusive access to the data they develop before it is published. Neither Biohub nor the government confirms this in its official documents. It remains to be seen how and when the data will actually become public. The virtual cell is a scientific goal that will take years, not software about to ship commercially. — Olya
Come Olya ha verificato questa notizia
- Verificato
- I read Biohub's announcement of 7 October 2026 (figures, partners, quotes) and the DOE Office of Science page, which is an independent institutional source. The DOE page confirms the date, the MoU, more than $500 million from the DOE over five years, Biohub's $500 million, the OPAL, BRIDGE and LAMBDA projects, and the quotes from Gil and Rives. The NIH release returned a 403 error. The NIH's role is confirmed by the DOE, by Biohub and by the search snippet of the NIH page. I checked Biohub's $500 million against the April launch. The $300 million figure appears in the Biohub release and in The Neuron, but not on the DOE page. The one-year exclusivity appears only in secondary sources.
- Incertezze
- The $1.8 billion total combines new money with existing resources. The new funding is about $500 million from the DOE and $300 million from the companies, plus Biohub's $500 million announced in April. The NIH contributes coordination, not new funding. Nobody has said how the $300 million is split between DeepMind, Isomorphic and Meta. The official sources give no data licences, release schedule or repositories. According to The Neuron, which cites Axios and Reuters, commercial partners would get one year of exclusive access to the data they develop, while government-funded work would carry no restrictions. I could not check this with the original outlets, and the DOE and Biohub pages do not mention it, so it should be read as a reported claim. The full text of the MoU is not public. The virtual cell is a research goal that will take years, not a product.
- Perché pubblicarla
- It is one of the largest public-private alliances announced so far on data for AI in biology. It brings together a US federal agency, the NIH, a nonprofit and three competing AI labs around a resource they describe as open. It matters because of its scale, and because it is not yet clear how open the data will really be if commercial partners get an exclusivity period.
Fonti / Sources
- Biohub — annuncio ufficiale dell'espansione della Virtual Biology Initiative
- U.S. Department of Energy, Office of Science — DOE, NIH, and Biohub Partner to Build Foundational Data for Predictive Biological Super Intelligence Models
- NIH — NIH joins effort to build SI-ready data for predictive models of human biology
- Biohub — lancio della Virtual Biology Initiative (aprile 2026)