Our body's tissue is like a city: different neighborhoods are home to different cells, each with its own job. To understand how this city is organized, biologists take images showing which genes are active at each spot. But sharp images are expensive, and cheap ones come out blurry, like a photo of a city from high above, where individual buildings are hard to make out.
The new method SRast learns to turn such blurry images into detailed ones. It works like a translator who knows many languages: first, it studies various tissues, from brain to heart, and remembers what their structures typically look like. Then, when it sees a new blurry image, it fills in the details without confusing one thing for another. Crucially, the method ensures that the amount of "substance" at each spot stays the same: if a coarse image shows little signal in a certain place, the detailed version must not invent extra signal. It's like zooming into a photo without being allowed to add buildings that weren't in the original.
Tests across different animal species and different instruments showed that the method recovers fine structures better than previous approaches. This paves the way for cheap and accurate tissue analysis in hospitals and labs, where one previously had to choose between cost and quality.
Spatial transcriptomics is a way to see which genes are active at each point in a tissue without destroying it. Previously, this required cutting the tissue into pieces, but now you can get a complete map.