If you want to know how researchers find new peptide hormones, the short answer is that they start with a computer, not a test tube. They search human genes for the places where an enzyme would cut a larger protein and release a short peptide, then test the best candidates on lab-grown cells and in animals. The clearest recent example is a Nature study from Stanford Medicine’s team, which used this approach to find a peptide called BRP (BRINP2-related peptide). You can read the Stanford Medicine’s release on BRP, reproduced by ScienceDaily, for the original wording.
The method is worth understanding step by step, because each step narrows the list and shows where the approach is strong and where it stops.
Step one: search human genes for cut sites
A prohormone is a larger protein that holds a hormone until an enzyme cuts it free. The enzyme that matters here is prohormone convertase 1/3, which cuts prohormones at specific sites. Each cut can release a shorter peptide.
The team built an algorithm called Peptide Predictor and pointed it at about 20,000 human genes. Its job was to find the sites this enzyme cuts. That gives you a list of possible peptides before anyone makes a single one in the lab, which saves time and materials.
How 20,000 genes became one candidate
Each filter shrank the list. Here is the funnel the team reported.
| Stage | Number | What happened |
|---|---|---|
| Human genes searched | About 20,000 | Searched for sites cut by prohormone convertase 1/3 |
| Prohormones left after filters | 373 | Proteins that could yield peptides |
| Possible peptides | 2,683 | Candidate peptides from those prohormones |
| Tested on lab-grown nerve cells | 100 | Included GLP-1 as a comparison |
| Standout | 1 | BRP, which raised neuronal activity tenfold |
Notice how the numbers fall at each stage. A computer handles the first two steps, and the lab takes over only for the shortlist of 100. That is the core of the method: cheap screening first, expensive testing last.
What the cell and animal tests showed
In lab-grown nerve cells, GLP-1 raised neuronal activity threefold. BRP raised it tenfold, and that comparison is what made BRP stand out from the other 99 peptides tested.
The team then moved to animals. Here is what they reported:
| Test | What the team reported |
|---|---|
| Food intake in lean mice and minipigs given BRP by injection | Cut by up to 50% over the following hour |
| Obese mice given injections for 14 days | Lost about 3 grams on average, while untreated mice gained about 3 grams |
| Body composition in the obese mice | Nearly all the weight lost was fat, not muscle |
| Glucose and insulin tolerance | Improved |
| Movement, water intake, anxiety-related behavior and fecal output | No significant differences from controls |
The team’s own explanation is that BRP appears to act in the hypothalamus, a brain region, rather than across the gut, pancreas and brain receptors that semaglutide targets. That is a hypothesis about where the peptide works, and the team describes it as such.
What a computer search cannot tell you
A computer search can rank candidates, but it cannot show you what they do in a living body. The team is direct about the gaps:
- It is not known whether people would see comparable effects.
- Nausea cannot be judged from animals.
- The receptors that recognize BRP have not yet been identified.
- Peptides break down quickly, so a longer-lasting version may be needed.
- Clinical trials must establish safety, effectiveness and dose.
Prediction-first screening is a general pattern in drug research, not something unique to peptides. A 2024 review of computational prediction methods, indexed by the U.S. National Library of Medicine, describes computer tools that predict which molecules might form a stable structure before any lab work begins, saving time and resources. That review concerns pharmaceutical cocrystals rather than peptides, but the logic is the same: predict first, test later.
The slow stage comes after the lab. The National Eye Institute’s 2004 uveitis report shows the same order with a different drug: laboratory studies in an experimental model came first, and a preliminary clinical trial in people followed.
Who funded the work and who holds patents
The study was funded by the NIH, Stanford programs, the American Heart Association and the Carlsberg Foundation. Co-researchers came from UC Berkeley, the University of Minnesota and the University of British Columbia. Senior author Katrin Svensson, PhD, is an assistant professor of pathology, and research was led by Laetitia Coassolo, PhD.
Svensson and Coassolo are named inventors on patents on BRP peptides for metabolic disorders. Svensson also cofounded Merrifield Therapeutics to advance BRP toward human trials. Knowing this helps you weigh the work, which the team itself says must be confirmed in clinical trials.
Where to read next
For the broader picture, start with exploring peptides, which covers the main types and uses. Then read the peptide research evidence guide to see how human studies and safety data compare with lab and animal results.
Disclosure: this article explains a research finding. The Stanford release describes a research finding and names no product, approval or seller, and this page does not recommend any product.
By the Clinical Research Explained Editorial Team