AI-driven cell classification company Deepcell announced it will partner with Stanford University to share data with Tabula Sapiens, a program with the goal of creating a benchmark human cell atlas of 2 million cells collected from 25 organs of eight people.
Tabula Sapiens is a benchmark, first-draft human cell atlas of nearly 500,000 cells from 24 organs of 15 normal human subjects. The work is a product of the Tabula Sapiens Consortium, funded by the Chan Zuckerberg Initiative’s single-cell biology program and the Chan Zuckerberg Biohub. The CZ Biohub is co-led by noted researchers Joe DeRisi of the University of California, San Francisco, and Stephen Quake, of Stanford University.
“We are excited to partner with Dr. Stephen Quake, one of the pioneers in the single-cell revolution at Stanford, to advance the understanding of cell biology,” said Maddison Masaeli, CEO of Deepcell in a press release. “The novel insights from our technology will be of particularly high value for multi-omic studies such as the Chan Zuckerberg Biohub’s Tabula Sapiens program. We believe that adding an entirely new dimension to understand cells through AI-powered cell morphology and linking it to whole transcriptome analysis will usher in a new era of discoveries in cell biology.”
The Tabula Sapiens program aims to take the organs from the same individual controls for genetic background, age, environment, and epigenetic effects to allow the detailed analysis and comparison of cell types that are shared between tissues. The intent is to create detailed portraits of cell types and better understand their distribution and variation across tissues and within the endothelial, epithelial, stromal and immune compartments.
As part of the collaboration, researchers will use Deepcell’s technology platform to generate single-cell morphology data. These data will then be made publicly available in Tabula Sapiens to be shared with the broader research community. This work will also help expand Deepcell’s Deep Cell Atlas which currently includes more than 1 billion cell images. The company’s AI-driven research platform has been trained on millions of images to identify cells based morphological patterns that may not be visible to the human eye.