California-based Biohub, the US Department of Energy, the National Institutes of Health, Google DeepMind, Isomorphic Labs, Meta, NVIDIA, and major research institutes have launched a nearly USD 1.8 billion effort to generate open, AI-ready biological datasets for models designed to predict how living cells behave in health and disease.
The Virtual Biology Initiative is anchored by a USD 500 million commitment from Biohub, a nonprofit biomedical research organization founded by Priscilla Chan and Mark Zuckerberg. That total includes USD 400 million for new measurement technologies such as cryo-electron tomography, high-throughput tissue microscopy, and molecular engineering tools, and USD 100 million for external research. The DOE will contribute more than USD 500 million over five years through its Genesis Mission, providing access to exascale computing and autonomous laboratory infrastructure across the US National Laboratory system.
NIH will contribute datasets, repositories, and knowledge bases built through more than USD 500 million in prior federal investment, while Biohub will work with the agency to standardize those resources for AI model training. Google DeepMind, Isomorphic Labs, and Meta are collectively investing USD 300 million to develop technologies and multimodal datasets, while NVIDIA is contributing accelerated computing infrastructure, software, and technical expertise.
The initiative aims to generate cell-response data across a broader range of cell types, states, and perturbations than has previously been available, with the long-term goal of building predictive “virtual cell” models that can simulate biological responses and potentially inform disease research and drug development.