The AI-Discovered Peptide That Curbs Appetite Without the Downsides

Key takeaways

• Stanford Medicine researchers used an AI algorithm to sift through the entire human genome and identify a naturally occurring 12-amino-acid peptide, called BRP, that suppresses appetite.

• In animal studies, BRP reduced food intake by up to 50% and led to fat-specific weight loss, while avoiding the digestive side effects and muscle loss associated with semaglutide (the active ingredient in many GLP-1 therapies).

• The key difference is precision: BRP appears to act narrowly on the brain's hunger-regulating region, rather than acting broadly across the gut, pancreas, and other tissues the way semaglutide does.

Finding a needle in a 20,000-gene haystack

Semaglutide's success came from copying a natural hormone, GLP-1, that regulates hunger and blood sugar. Stanford researchers wondered whether the same enzyme that helps produce GLP-1—prohormone convertase 1/3—might also be quietly producing other useful appetite-related signals hiding in the genome. The problem was scale: a single precursor protein can be cut into countless fragments, and only a small handful ever function as real hormones.

To solve this, the team built an algorithm called Peptide Predictor, which scanned all 20,000 human protein-coding genes for the specific cut sites this enzyme targets. That search narrowed 373 candidate precursor proteins down to a shortlist, and from there, the researchers tested 100 promising peptide fragments—including GLP-1 itself—on lab-grown neurons to see which ones activated the cells most strongly.

A tiny molecule with an outsized signal

GLP-1 performed as expected, tripling neuronal activity compared with untreated cells. But one far smaller peptide—just 12 amino acids long—did something more dramatic, increasing neuronal activity tenfold. The researchers named it BRP, after the parent protein it's cut from.

What happened next moved the discovery from a cell-culture curiosity to something with real physiological weight. In lean mice and minipigs (chosen because their metabolism closely resembles ours), a single injection of BRP before feeding cut food intake by as much as 50% within the following hour.

Weight loss without the usual tradeoffs

Researchers then gave daily BRP injections to mice carrying excess body fat over 14 days. The treated animals lost weight—almost entirely from fat—while animals in the control group gained weight over the same period. Treated mice also showed improved blood sugar and insulin response.

The more notable part came from what didn't happen. Behavioral testing showed no meaningful differences in movement, water intake, anxiety-like behavior, or digestive output between treated and untreated animals—no signs of the nausea or slowed digestion that often accompany semaglutide. That distinction traces back to where each molecule acts: semaglutide's target receptors show up throughout the gut, pancreas, and other tissues, which explains its wider range of effects, while BRP appears to act specifically in the hypothalamus, the brain region that governs hunger and energy use.

Why targeted beats broad

This is really a story about precision. Semaglutide works, in part, because it's broadly active—but that same breadth is what produces its side-effect profile. BRP's apparent selectivity for one specific brain region suggests it might be possible to get comparable appetite and weight effects while sidestepping some of the downstream tradeoffs that come from acting everywhere at once.

That said, this is still animal research, and important groundwork remains—identifying the receptor BRP binds to, mapping what happens after it binds, and figuring out how to make the peptide last longer in the body before it could reasonably move toward human trials.

The takeaway

This isn't a signal to expect a new option at the pharmacy anytime soon, but it is a good example of how AI-assisted discovery is opening up entirely new categories of naturally occurring molecules that regulate appetite and metabolism. If BRP's early promise holds up in further research, it points toward a future where managing body weight might not require accepting the same tradeoffs that come with today's most effective options.

References:

Coassolo, L., Danneskiold-Samsøe, N.B., Nguyen, Q., et al. Prohormone cleavage prediction uncovers a non-incretin anti-[obese] peptide. Nature, 2025; 641(8061): 192. DOI: 10.1038/s41586-025-08683-y.