UChicago Scientists Map Entire Mouse Body at Cellular Scale to Build Foundation for "Virtual Laboratory"

UChicago researchers develop an AI-powered spatial transcriptomics system to map every cell and organ in a mouse, building a foundation for a virtual laboratory.

By: AXL Media

Published: Mar 28, 2026, 10:59 AM EDT

Source: Information for this report was sourced from University of Chicago

A Continental Map of Biological Processes

While 21st-century biology has excelled at identifying gene expression within individual cells or isolated tissue samples, the ability to see how these processes interact across an entire body has remained an elusive goal. A research group led by Associate Professor Nicolas Chevrier at the University of Chicago has bridged this gap by developing a "molecular cartography" system. Published in the journal Cell on March 27, 2026, the study outlines a method to visualize how diseases and therapeutics affect every major organ and tissue region simultaneously. This holistic perspective is designed to move science away from fragmented observations and toward a unified understanding of systemic biology.

Engineering the Perfect Whole-Body Slice

The technical foundation of this breakthrough is the ability to prepare a specimen that is both massive in scale and microscopic in detail. Working with experts from Tsurumi University in Japan, the team developed a technique to create a single, intact cross-section of a frozen laboratory mouse that is the thickness of an average cell. Transferring such a delicate, large-scale sample onto a slide while preserving its genetic material required a significant optimization of "Array-seq" technology. This process allows researchers to capture high-resolution microscopy and genetic sequencing data from a single, body-wide snapshot rather than stitching together hundreds of smaller, disconnected images.

Artificial Intelligence as a Virtual Pathologist

To process the staggering amount of data generated by these whole-body sections, the team collaborated with AI experts at Fudan University to create a specialized machine learning model. Traditionally, identifying every cell type in a sample would require expensive and time-consuming chemical staining using various antibodies. The new AI model, however, can virtually label every organ and cell type using only standard hematoxylin and eosin staining—the most common and inexpensive diagnostic tool in medicine. This "virtual staining" allows researchers to annotate cellular information across the entire body at a fraction of the traditional cost and time.