The US government and major technology companies are joining a major initiative led by nonprofit Biohub to create large open datasets that can be used to train artificial intelligence models for biological research.
Biohub said on Wednesday that the combined commitments have brought the total investment in the project to about $1.8 billion, with the goal of accelerating scientific discoveries and potentially reducing the time required to develop new medicines.
Meta, Google DeepMind and drug discovery company Isomorphic Labs are jointly contributing $300 million to the initiative.
The US Department of Energy is also committing more than $500 million over five years to support laboratory measurements, modelling and computing, while the National Institutes of Health will coordinate datasets and repositories created through more than $500 million in previous federal funding.
Biohub itself committed $500 million to the project in April. The organisation was founded as a philanthropic venture by Meta CEO Mark Zuckerberg and his wife, Dr Priscilla Chan.
The initiative, known as the Virtual Biology Initiative, aims to generate large amounts of data showing how cells respond to different conditions. Researchers hope the information can be used to build predictive AI models capable of helping scientists understand biology and shorten drug development timelines that currently take years.
“Biology has been just sort of a clever discovery-based science until this point,” Chan said.
She added that the project was designed as a shared resource that could allow researchers to build on the work over time.
Datasets to Eventually Become Public
According to Biohub, the biological datasets generated through the initiative will eventually be made available to the public.
However, companies that provide commercial funding will receive temporary access before the data becomes publicly available.
Alex Rives, Biohub’s head of science, said the arrangement allows private investment to support a project that is ultimately intended to benefit the wider scientific community.
“With commercial funders we have embargo periods where there’s a period of time where the groups can work on the data, and then it becomes available as a public scientific resource,” Rives said.
He said government-funded research carried out alongside the privately funded work would not have the same restrictions.
Biohub also plans to seek additional support from pharmaceutical companies and philanthropic organisations.
Billions of Cells Needed for AI Models
Rives said existing biological datasets contain information from hundreds of millions of cells, but much larger quantities of data will be required to develop highly accurate predictive models.
According to him, the field will eventually need datasets covering billions and potentially trillions of cells.
“We need to capture the language of biology, we need to capture the language of the cell. And that doesn’t exist today,” he said.
The project will use technologies including spatial transcriptomics, which allows scientists to map molecular activity within intact tissue, as well as experiments that record how cells respond to changes in their surroundings.
Biohub said much of this information has never previously been generated in a coordinated manner.
Rives said work that would normally take decades is being targeted for completion within five years. The first dataset is expected to be ready in roughly a year, while researchers hope to have accurate predictive models within five years.
AI Companies Expand Biology Research
The Biohub initiative comes as other AI organisations are also increasing their investment in biological research.
Anthropic has expanded its biology work to include a wet laboratory, while the OpenAI Foundation has launched a grant programme worth more than $125 million to support biological and medical datasets for AI research.
The growing investment reflects increasing efforts to combine artificial intelligence with large-scale biological data, with researchers hoping the technology can improve understanding of diseases and accelerate the discovery of new treatments.
