Google, Meta and Isomorphic Labs back Biohub’s push to build a virtual cell

Google DeepMind, Meta and AI drug-discovery company Isomorphic Labs are putting $300 million into Biohub to help build the datasets needed for a “virtual cell,” a digital model that could let scientists test biological ideas on a computer before running lab experiments. The funding is part of a broader $1.8 billion effort that brings together tech companies, federal agencies and researchers in one of the most ambitious biology-and-AI collaborations to date.

The investment matters because it could accelerate how researchers study disease, design treatments and understand cells at a systems level. Biohub says the goal is to create data resources that allow scientists to ask, predict and answer biological questions digitally, potentially reducing the time and cost of discovery in medicine.

What is Biohub trying to build?

Biohub is working toward a scientific model that can simulate the behavior of a cell with enough accuracy to be useful for research. In practical terms, that means using AI datasets and computational methods to reproduce how genes, proteins, molecules and cellular processes interact inside living systems.

The organization, founded in 2016 by Mark Zuckerberg and Priscilla Chan, has framed the project as a way to create a “virtual cell” that scientists can use to run digital experiments. If successful, researchers could explore hypotheses in software first, then reserve costly physical experiments for the most promising leads.

Why does a virtual cell matter?

A virtual cell could make biological research faster, cheaper and more predictive. Today, many discoveries still depend on trial-and-error lab work, which can be slow and expensive. A reliable digital model would not replace laboratories, but it could narrow the search space and help scientists focus on the most relevant experiments.

Biohub’s science leadership says the broader aim is to create an accurate predictive model of biology that can speed discovery by letting scientists test ideas digitally before stepping into the lab. That kind of capability could have major implications for disease research, drug development and personalized medicine.

Who is putting money into the project?

The new funding round includes three private-sector names that carry outsized weight in AI: Google DeepMind, Meta and Isomorphic Labs. Together they are contributing $300 million to Biohub, according to reporting from Reuters.

Those companies are not alone. The project is also receiving support from public institutions, including the US Department of Energy and the National Institutes of Health, signaling that the effort is being treated as a national-scale science project rather than a standard corporate partnership.

How will the money be used?

The money will go toward building the datasets and scientific infrastructure needed to make the virtual-cell project viable. Biohub says the effort is designed to create high-quality AI datasets that can help researchers model biological systems more effectively.

Those datasets are central to the project because biology is messy, multi-layered and difficult for AI systems to learn from without large, reliable inputs. The more comprehensive and well-structured the data, the more useful the resulting models are likely to be for researchers.

What kind of data is Biohub seeking?

Biohub is focused on creating the kinds of datasets that can support digital biology at scale. That includes information from existing research repositories as well as new data generated through coordinated efforts across institutions.

The NIH’s role is especially important here. The agency is contributing datasets, repositories and knowledge bases that stem from earlier federal investment, giving the project a foundation of existing scientific material rather than requiring everything to be built from scratch.

Why are Google, Meta and Isomorphic Labs involved?

They each bring different strengths to a problem that sits at the intersection of AI, life sciences and large-scale computing. Google DeepMind has deep experience in advanced machine learning research. Meta has invested heavily in AI infrastructure and open scientific research. Isomorphic Labs, meanwhile, is focused on AI-driven drug discovery.

That combination suggests the project is not only about funding basic science but also about building reusable AI tools that could influence future biomedical research and drug development. For the companies involved, the partnership also offers a chance to shape the next generation of biological modeling.

Biohub’s science chief, Alex Rives, said a truly accurate predictive model of biology could significantly speed up discovery because scientists would be able to test ideas digitally. He also said the virtual-cell project is one of the most important scientific challenges of the coming era and will require large-scale coordinated data generation across institutions and countries.

How big is the broader initiative?

The $300 million private investment is only one slice of a much larger $1.8 billion program. That scale matters because it shows the initiative is aiming well beyond a single tool or research pilot. Instead, it is trying to assemble the data, compute and institutional backing required for a long-term scientific platform.

Public-sector involvement gives the project additional legitimacy and access to decades of biomedical knowledge, while the private companies add AI expertise and technical resources. Together, the partners appear to be trying to solve a problem that no single organization is likely to crack alone.

Timeline of the Biohub effort

What challenges stand in the way?

Building a virtual cell is a formidable scientific and engineering problem. Cells are extraordinarily complex systems, and biology does not always behave in neat, predictable ways that machines can easily learn from. Even with better data and more powerful models, researchers will still have to deal with incomplete knowledge, noisy measurements and the challenge of validating predictions in the real world.

There is also the question of scale. To be genuinely useful, a virtual cell would need to reflect enough of biological reality to help answer meaningful questions without oversimplifying the processes it is meant to represent. That means the project will depend not only on computing power, but also on the quality of the underlying science and coordination across many institutions.

Why coordination matters

Biohub’s leadership says the project requires data generation at national and international scale, underscoring that no single lab or company can build something this ambitious on its own. The promise of the initiative depends on shared standards, interoperable datasets and sustained collaboration across public and private sectors.

That coordination challenge is one reason the project has attracted such a broad coalition. In fields like genomics and drug discovery, the most valuable resources are often the ones that are hardest to standardize. A shared framework could help researchers move from scattered studies to integrated biological models.

What could this mean for medicine and drug discovery?

If the project succeeds, it could change how scientists approach disease prevention and treatment. Researchers might use virtual simulations to identify drug targets, understand how cells respond to different interventions or compare how biological systems behave under different conditions.

In drug discovery, that could mean fewer dead ends and more targeted experiments. In medicine, it could help researchers uncover patterns that are difficult to spot using conventional methods alone. The potential benefits are substantial, but they are still prospective: Biohub is building the scientific foundation, not announcing a finished tool.

• It could reduce the number of costly laboratory experiments needed at early stages.
• It may help researchers test biological hypotheses more quickly.
• It could improve disease modeling and treatment discovery over time.
• It may establish shared AI datasets for future biomedical work.

How this fits the current AI race

The Biohub investment also reflects a broader shift in the AI industry: the move from generic chatbots and consumer tools toward science-heavy applications with real-world stakes. Tech giants increasingly see biology, chemistry and medicine as fertile ground for AI systems that can do more than generate text or images.

By backing a virtual-cell initiative, Google DeepMind and Meta are placing a bet on AI as infrastructure for scientific discovery, not just a product layer. That position could prove strategically important if biology becomes one of the next major frontiers for advanced AI.

For Isomorphic Labs, the partnership aligns with its core mission of using AI to accelerate drug discovery. For Biohub, it provides the technical and financial support needed to pursue a challenge that could take years to mature.

What happens next?

The next phase will likely focus on dataset creation, model development and coordination among the participating institutions. Because the project is so dependent on scientific infrastructure, progress may be incremental rather than dramatic, with early gains coming from better data quality and more standardized research inputs.

Still, the scale of the investment is a signal that major players believe the idea is worth pursuing now. If the virtual-cell concept advances, it could become one of the clearest examples of AI moving from digital conversation into the core machinery of science.

For now, the headline is not that a virtual cell already exists. It is that some of the biggest names in tech and biomedical research are committing serious money and institutional backing to make one possible.