A major new scientific push aims to transform medicine by teaching artificial intelligence to understand how human cells function. Biohub has announced the Virtual Biology Initiative, a five-year global program backed by a $500 million commitment to generate the data and technologies needed for predictive AI models of life.
The project seeks to create large-scale open datasets that could allow researchers to digitally simulate cellular behavior, helping scientists uncover disease mechanisms faster and accelerate the discovery of new treatments. Biohub said the initiative will make all generated data openly available to the worldwide research community.
Of the total investment, $100 million will support external research and international data-generation efforts, while $400 million will fund next-generation imaging, engineering tools, computing infrastructure, and internal large-scale data production.
Several leading institutions are already joining the effort, including Allen Institute, Arc Institute, Broad Institute, Wellcome Sanger Institute, the Human Cell Atlas, and the Human Protein Atlas. NVIDIA will also provide accelerated computing systems and technical expertise.
Scientists believe accurate predictive cell models could revolutionize biomedical research by allowing experiments to be performed digitally at a speed and scale impossible in traditional laboratories. This could lead to faster breakthroughs in cancer, neurological diseases, immune disorders, and many other complex conditions.
According to Biohub’s Head of Science Alex Rives, the challenge now is data. Existing biological datasets remain too limited to train advanced AI systems capable of representing the full complexity of living cells.
