The Chan Zuckerberg Initiative (CZI) has announced an ambitious new effort to build one of the world's largest computing systems dedicated to nonprofit life sciences research. The system, which will be equipped with over 1,000 GPUs, represents a major investment in harnessing the power of artificial intelligence to accelerate discoveries in biology and medicine.
In an essay published in MIT Technology Review, CZI co-founders Priscilla Chan and Mark Zuckerberg elaborated on the motivations behind this computing cluster. They envision researchers using the system to create advanced "virtual cell" models capable of predicting how real cells behave in health and disease.
"Scientists could use such a simulator to predict how cells might respond to specific conditions and stimuli: how an immune cell responds to an infection, what happens at the cellular level when a child is born with a rare disease, or even how a patient’s body will respond to a new medication," the co-founders wrote. By computationally modeling cell systems, scientists may gain profound new insights into cell biology and disease mechanisms.
The CZI-funded system will provide nonprofit researchers with access to capabilities that have thus far been prohibitively expensive for many in academia. The initiative aims to democratize access to resources needed to apply large language models and other cutting-edge AI techniques to open bioinformatics datasets.
CZI's approach to cell modeling is two-fold: the hardware, represented by the computing cluster, and the software, symbolized by the immense datasets that will power these models.
The computing system will train on diverse cellular data aggregated by CZI and partners, encompassing tens of millions of cells. These include comprehensive maps of cell types from the Human Cell Atlas project and subcellular protein atlases from CZI's imaging and biohub researchers.
Combined with CZI's existing tools like CELL by GENE (CZ CELLxGENE) for visualizing single-cell data, the new cluster represents an ambitious vision of collaborative science powered by AI. According to Chan and Zuckerberg, "With a new AI cluster trained on this data, we believe researchers’ rate of discoveries could significantly increase."
The goal is to develop advanced in silico models capable of predicting cellular responses to diseases, medications, and other perturbations. As described by CZI Co-CEO Priscilla Chan, “AI models could predict how an immune cell responds to an infection, what happens at the cellular level when a child is born with a rare disease, or even how a patient’s body will respond to a new medication.”
While acknowledging the scientific challenges ahead, CZI's leaders believe the initiative could catalyze breakthroughs leading to new therapies or even cures for intractable diseases. They see the effort as emblematic of CZI's mission to marshal data, technology, and collaboration to overcome daunting health challenges.
By pooling resources to create open AI models demystifying the fundamental workings of cells, Chan and Zuckerberg hope to propel biomedicine into a new era of insight. While substantial hurdles remain, CZI's computing investment highlights the transformative potential of aligning AI with life science in service of understanding and improving human health.