Biohub, DOE, and NIH Commit $1.8B to Virtual Biology Initiative for AI-Ready Biological Data

Biohub, the U.S. Department of Energy (DOE), and the National Institutes of Health (NIH) have announced a major expansion of the Virtual Biology Initiative, committing a total of $1.8 billion in funding, data, computation, and new measurement technology to generate open, AI-ready biological data. This coordinated effort aims to build foundational datasets that enable predictive AI models to simulate cellular responses, accelerating the discovery of treatments for human diseases.

Funding and Partnerships

The initiative aggregates resources from public and private sectors to scale data generation beyond the capacity of any single organization:

• Biohub: Anchors the effort with a $500 million founding commitment. $400 million funds new measurement technologies, including cryo-electron tomography and high-throughput microscopy, while $100 million supports external research.
• Department of Energy (DOE): Investing more than $500 million over five years through the Genesis Mission. This funding supports lab measurement, modeling, and computation using exascale supercomputers and automated laboratories.
• National Institutes of Health (NIH): Contributing via the Bio Genesis Mission by coordinating the contribution of relevant datasets, repositories, and knowledge bases developed through more than $500 million in prior federal investment aligned to this initiative.
• Private Sector: Google DeepMind, Isomorphic Labs, and Meta are collectively investing $300 million to develop the necessary technologies and multi-modal datasets.

Technical Scope and Goals

The Virtual Biology Initiative seeks to create a "virtual cell" capable of predicting how biological systems respond to perturbations. Key technical objectives include:

• Data Scale: Expanding measurements to cover millions to billions of cells across diverse conditions, far exceeding current datasets.
• Technology Integration: Combining cryo-electron tomography for subcellular resolution with live-cell imaging and engineering tools to manipulate cells at tissue and organism levels.
• Open Access: The resulting datasets will be open resources for the global scientific community, intended to enable researchers to train models that can simulate disease progression and test therapeutic interventions in silico.

Institutional Support

Leading scientific institutions, including the Broad Institute, Human Cell Atlas, and Wellcome Sanger Institute, are joining the collaboration to help standardize data and define scientific priorities. NVIDIA is also involved, providing AI infrastructure and software to support the computational demands of modeling complex biological systems.