Five of the most-studied research peptides: an evidence map

This is an evidence map, not a ranking. It takes five peptides with substantial scientific literatures and asks the same three questions of each: what research areas they have been studied in, what the strongest type of evidence is, and where that evidence stops. The compounds appear in alphabetical order. No compound is presented as better than another, and nothing here is a recommendation.

How the five were chosen

We counted PubMed records for each peptide covered in this library, searching titles and abstracts in September 2026. The five with the largest literatures were selected, taking one representative where two compounds belong to the same drug class. Semaglutide therefore stands for the incretin-based medicines, and tirzepatide, covered in its own article, was not included separately. Melanotan II was also left out. Its abbreviation overlaps with unrelated terms in searches, which made a reliable count impossible, and much of its literature uses it as a laboratory tool in animal studies. It is covered in a separate article.

These figures are approximate. They depend on search terms and indexing, and they change over time. A study indexed as a randomised trial may measure a peptide rather than administer it. That is true of several MOTS-c exercise studies, for example.

Why publication counts mislead

A large literature can look like strong evidence. It is not the same thing. Counts include reviews, commentaries and repeated studies from the same laboratories. They say nothing about study size, design or independent replication. Positive results are also more likely to be published than negative ones. In animal research on stroke, publication bias was estimated to account for around a third of the reported efficacy^[1]. BPC-157, GHK-Cu and MOTS-c illustrate the point. Their literatures are of similar size. Yet GHK-Cu has a controlled human trial, MOTS-c has large human genetic studies, and BPC-157 has neither.

The map at a glance

BPC-157

• Research domain. Healing in the gastrointestinal tract, tendon, ligament, muscle and bone, mainly in rats.
• Strongest evidence. Animal studies. A 2025 systematic review found 35 preclinical studies and one clinical study among its 36 eligible studies^[2].
• Major limitations. Human data are limited to three small pilot studies without rigorous trials^[3]. The largest is a retrospective review of 16 patients with no control group^[4]. Most of the literature comes from a single research group, as our separate analysis of the evidence base discusses.

GHK-Cu

• Research domain. Collagen and extracellular matrix biology, skin, wound healing and, more recently, gene expression.
• Strongest evidence. Laboratory studies show that the copper complex stimulates collagen synthesis in cultured fibroblasts at very low concentrations^[5]. The most important human study is a multicentre, randomised, evaluator-blinded, placebo-controlled trial in diabetic foot ulcers. It reported a median closure of 98.5% of ulcer area with GHK-Cu gel, against 60.8% with vehicle^[6].
• Major limitations. That trial dates from 1994 and was not replicated in the sources reviewed. The broadest claims come from gene-expression analyses. One of these identified GHK computationally as a compound able to reverse an emphysema-related expression signature^[7], which is a hypothesis rather than a clinical finding. Reviews summarising cosmetic benefits^[8] rely on data that are hard to evaluate independently.

MOTS-c

• Research domain. Metabolism, insulin sensitivity, exercise capacity and ageing.
• Strongest evidence. Mouse studies. The discovery paper showed that MOTS-c is encoded in mitochondrial DNA and acts mainly on skeletal muscle. In mice, MOTS-c treatment prevented insulin resistance related to age and to a high-fat diet, as well as diet-induced obesity^[9]. Later work reported improved physical capacity in old mice^[10].
• Human evidence. Human data are observational. A genetic variant that alters MOTS-c was associated with type 2 diabetes in men in a meta-analysis of 27,527 people^[11]. Exercise studies have measured circulating MOTS-c in blood^[12].
• Major limitations. No controlled trials of giving MOTS-c to people were identified in the sources reviewed.

Semaglutide

• Research domain. Type 2 diabetes, obesity and cardiovascular risk.
• Strongest evidence. Large randomised, double-blind trials. In STEP 1, 1,961 adults with obesity lost 14.9% of body weight over 68 weeks with semaglutide, against 2.4% with placebo^[13]. In SELECT, 17,604 people with cardiovascular disease and overweight or obesity, without diabetes, had fewer major cardiovascular events with semaglutide than with placebo (hazard ratio 0.80)^[14].
• Major limitations. The pivotal trials involved the manufacturer. The evidence applies to a licensed pharmaceutical product used within trial conditions. It cannot be transferred to other sources or formulations.

Thymosin β4

• Research domain. Wound healing and repair in the skin, eye, heart and brain^[15].
• Strongest evidence. Animal studies of tissue repair, together with small randomised human trials. A placebo-controlled phase 1 study in 40 healthy volunteers found intravenous thymosin β4 well tolerated^[16]. A phase 2 trial in nine patients with severe dry eye reported improvements in discomfort and corneal staining with thymosin β4 eye drops^[17].
• Major limitations. The human trials are few and small. The published report of a venous ulcer trial describes its design and early enrolment rather than its results^[18]. None of this evidence concerns TB-500, a short fragment that has barely been studied in its own right.

Reading the map

Arranged by evidence type rather than reputation, the five compounds fall into three broad groups:

1. Semaglutide. Supported by large randomised trials with clinical outcomes.
2. Thymosin β4 and GHK-Cu. Substantial laboratory and animal work, with a few small controlled human trials that have not been scaled up.
3. MOTS-c and BPC-157. Evidence that is predominantly animal or observational, with no controlled human trials of administration.

The difference between these groups is qualitative, not a matter of degree. A mouse study, however well conducted, answers a different question from a randomised human trial. Adding more mouse studies does not turn one into the other.

The map is not fixed. The position of any compound would change with new evidence of the right kind. That means independent replication of key animal findings, controlled human trials with predefined clinical outcomes, and full publication of results whether positive or negative. For the less-studied compounds here, a single well-designed randomised trial would add more to the evidence base than many further studies of the same design.

Summary

The five peptides with the largest literatures in this library have very different kinds of evidence. Semaglutide is supported by large randomised trials, including a cardiovascular outcome trial. Thymosin β4 and GHK-Cu have small human trials built on laboratory and animal research. MOTS-c and BPC-157 rest mainly on animal and observational data. The size of a literature says little about its strength. What matters is the type of study, its design and independence, and whether findings have been replicated in people.