Mice in space grew brown fat that burns calories — first AI analysis of NASA's space biology data

Female mice exposed to 37 days of microgravity in orbit showed a 12-fold increase in Ucp1, a gene that controls thermogenesis (heat-burning metabolism), compared to ground controls. Researchers used machine learning to identify which genes predicted this shift, finding that the thermogenesis pathway activates as a compensatory response to weightlessness.

This is the first application of explainable machine learning to a NASA space biology dataset, which means the Open Science Data Repository now has a reusable model that other researchers can apply to existing microgravity experiments. The concrete finding—that female mice's fat tissue reprograms toward heat production in weightlessness—matters because astronauts on long-duration missions experience similar metabolic stress, and understanding the mechanism could inform countermeasures. It also opens a path: space biology datasets that were previously hard to extract insight from can now be re-analyzed with modern ML tools, potentially revealing patterns that older statistical methods missed.