Stanford finds appetite peptide with AI
A Stanford-led team used AI to identify a peptide called BRP that reduced appetite and body weight in animals and may work like Ozempic but without some side effects. The result is presented as early-stage research rather than a drug approval, and the report highlights the discovery role AI played in the finding ((sciencedaily.com)).
Many hormones start as larger proteins that cells cut into smaller working pieces, like slicing a long strip into useful parts. A Stanford-led team used artificial intelligence to search those hidden fragments and found one called BRINP2-related peptide, or BRP, that curbed appetite in animals. (pubmed.ncbi.nlm.nih.gov) The study was published March 5, 2025, in *Nature* by researchers at Stanford University, the University of California, Berkeley, the University of Minnesota, and the University of British Columbia. Senior author Katrin Svensson and lead author Laetitia Coassolo said the search focused on peptides produced by an enzyme called prohormone convertase 1/3, or PCSK1, which has links to human obesity. (pubmed.ncbi.nlm.nih.gov; nature.com) Artificial intelligence did not test the drug in animals; it narrowed the hunt. The researchers reported that their computational approach mapped more than 2,600 previously uncharacterized human peptide fragments and singled out the 12-amino-acid BRP for follow-up experiments. (pubmed.ncbi.nlm.nih.gov) BRP reduced food intake and showed anti-obesity effects in mice and pigs, according to the paper’s abstract. Stanford Medicine said the animal work also suggested less nausea, less constipation, less food aversion, and less muscle loss than semaglutide, the medicine sold as Ozempic and Wegovy. (pubmed.ncbi.nlm.nih.gov; med.stanford.edu) Semaglutide copies the action of glucagon-like peptide 1, or GLP-1, a hormone involved in appetite and blood sugar control. Stanford said semaglutide’s targets are found in the brain, gut, pancreas, and other tissues, while BRP appeared to act more specifically in the hypothalamus, a brain region that helps control hunger and metabolism. (med.stanford.edu) That difference in where the signal lands is the main reason researchers are paying attention to BRP. In a *Nature Reviews Endocrinology* research highlight, outside commentators said computational tools that predict enzyme-cut peptide sequences can speed the discovery of bioactive molecules with therapeutic potential. (nature.com) The researchers also checked whether BRP exists naturally rather than only as a lab design. The review said they detected BRP in human cerebrospinal fluid and in mouse brain tissue, and found lower brain levels in mice with impaired PCSK1 function. (nature.com) This is still preclinical research, not an approved treatment for obesity. Stanford said Svensson has co-founded a company to start human clinical trials “in the near future,” but no human efficacy data were reported in the March 2025 paper. (med.stanford.edu; pubmed.ncbi.nlm.nih.gov) So the immediate story is less that a new obesity drug has arrived than that a new way to find one may be working. In this case, the algorithm found a naturally occurring peptide that standard lab searches had missed, and the animal data were strong enough to move the idea toward human testing. (pubmed.ncbi.nlm.nih.gov; med.stanford.edu)
