Scientists map whole-body biomolecular models

Biology usually gets studied one slice at a time. One organ. One tissue. One pathway. But disease does not respect those boundaries — inflammation in the gut can change the liver, the brain, the immune system, and everything in between. That is why this new whole-body mapping work matters. A team at the University of Chicago just pushed spatial biology out from tiny tissue windows to something much bigger: gene-expression maps across entire mouse body sections, with enough detail to connect organs, tissues, and many cell types in one view. ### What actually got built? The new platform combines specimen preparation, spatial transcriptomics, and machine learning to read gene activity across whole mouse body sections instead of isolated tissue samples. Spatial transcriptomics already lets researchers see where genes are active inside tissue, but the usual tradeoff is scale — you get a beautiful local map, not an organism-wide one. This system stretches that idea across the body. ### Why is whole-body the hard version? Because tissues are wildly different from each other. The liver is dense and metabolically busy. Lung tissue is airy. Bone is hard. Brain tissue has its own structure entirely. A method that works on one organ often breaks on another. The Chicago team’s advance was not just collecting more data — it was making a pipeline that could handle many organs in one body-wide experiment and then align the results computationally. ### How much did they actually map? A lot, but not “the whole body in full detail” yet. The team says the system mapped all organs and tissue regions in the mouse sections they studied and identified about 75% of known mouse cell types. That makes this more like a serious operating system for whole-body biology than a finished digital twin. It is broad enough to study systemic responses, but still incomplete at the cell-type level. ### What did they use it for first? They used it to look at systemic inflammation in endotoxemia — basically a model of body-wide inflammatory stress. That matters because inflammation is exactly the kind of process that refuses to stay local. Instead of asking what one organ did, the platform let the researchers trace gene-expression changes across tissues and cell types throughout disconnected lab assays. ### Is this the same as a virtual human? Not yet. Think of this as infrastructure. Projects like HuBMAP are building a 3D reference atlas of the healthy human body, with 4,499 anatomical structures, 1,195 cell types, and 2,089 biomarkers in its 2025 release. Other AI systems, like Harvard’s PINNACLE, are getting better at modeling protein behavior in specific cellular and tissue contexts. The new mouse work is a predictive human simulator. ### So why are people excited? Because drug responses, immune reactions, and disease cascades are whole-body problems. If you want to predict toxicity, transplantation responses, or why a therapy helps one organ but harms another, single-organ maps are not enough. Reviews and perspective pieces over the past year have been converging on the same point: AI is becoming useful when it links molecules to cells, tissues, organs, and system-level outcomes in one framework. ### What is still missing? Human data at this level, for one thing. Also time. A real predictive model has to show not just where molecules are, but how states change hour by hour under disease, treatment, and recovery. And then there is causality — a map can show correlations across organs, but turning that into trustworthy intervention predictions is much harder. The catch is that “whole-body model” can mean anything — the field is still climbing that ladder. ### Bottom line The news is not that scientists built a full virtual human. They did not. The real step is narrower and more important: they showed that body-wide molecular mapping can now be done at useful scale in mice. That gives researchers a better shot at building models that treat biology as one connected system — which, turns out, is how disease has been behaving all along.