Research Peptide Fundamentals research peptides, and the trials that would settle them

In plain terms

A study can only answer the question it was built to answer. A test in a dish shows that a molecule does something to cells. A test in mice shows that it does something to a mouse. Neither one shows that it helps a person. Learning that requires a trial in people, with a group that does not get the compound, and a result worth measuring.

This site lines up four research peptides and asks one question about each. Does the study that would settle the popular claim about it actually exist?

For one of the four, tirzepatide, it does. For BPC-157, MOTS-c and CJC-1295 it does not, and each is short a different study.

So these pages do two things. They report what has been measured, with the source attached to the number. Then they describe the study that has not been run, and what it would have to look like to count.

Question first, design second

Most peptide reading runs backwards. A result arrives first — a percentage, a fold change, a headline — and the design that produced it gets checked later, if at all. Reversed, the exercise becomes tractable. Name the question. Name the design that can answer it. Then look for that design in the record.

Each class of question has exactly one family of designs that can settle it.

• Does the molecule touch a target at all? A cell-free binding assay settles that, and nothing else.
• Does it change a process in a living organism? An animal model settles that, in that species.
• Does the molecule track with anything in people? An observational cohort settles that, and it establishes association only.
• What does it do to the body's chemistry, and is it tolerated? An open-label pharmacology study in volunteers settles that.
• Does giving it change an outcome? Only a randomised trial with a control group settles that.
• Is it better than the existing option, and does the benefit hold? Only an active-comparator trial and a long follow-up settle that.

A design cannot be promoted. A rodent experiment run larger is still a rodent experiment, and a cohort study run longer still cannot show that an intervention caused anything. Those six rows are the rows of the matrix. Each of the four compounds fills a different subset of them, and the empty cells are where this planner does its work.

What a research peptide is

A peptide is a short chain of amino acids — the same building blocks as a protein, in a chain short enough to synthesise directly. Length is a practical dividing line rather than a formal one. BPC-157 runs fifteen residues. MOTS-c runs sixteen. Tirzepatide runs thirty-nine.

"Research peptide" is a regulatory description, not a pharmacological one. It marks a compound distributed for laboratory work rather than approved for people, and three of the four here carry that status. BPC-157 is not approved as a medicine anywhere and moves through non-regulated channels, so identity, purity and content are unverified outside formal studies ^[2]. MOTS-c has no approved indication, formulation or human dosing standard of any kind. CJC-1295 was reviewed at the 2024 FDA Pharmacy Compounding Advisory Committee and was not recommended for the compounding bulk substances list.

Tirzepatide (Mounjaro, Zepbound) is the exception, and it earns its place here for exactly that reason. It is an approved prescription drug, cleared in May 2022 for type 2 diabetes ^[14], which is why its evidence base looks nothing like the other three. Set beside them it functions as a scale: this is what the finished article looks like, and this is how far the others are from it.

Four compounds, four different missing studies

The four were not selected because they resemble each other. They were selected because their records break in four different places.

BPC-157 has volume without translation. A large rodent literature, an angiogenic mechanism running through the VEGFR2 receptor with downstream Akt and nitric-oxide signalling ^[4], and a foundational result showing reduced gastric ulcer area in rats ^[5]. In people: three small pilot reports, one of which gave intravenous BPC-157 to two healthy adults and recorded no adverse events and no measurable change in cardiac, hepatic, renal, thyroid or glucose markers ^[1]. A 2025 narrative review states plainly that rigorous large-scale trials are lacking and that the compound should be treated as investigational ^[2]. Nothing in that record was designed to test whether a human tendon heals faster.

MOTS-c has human data of the wrong kind. It is the only one of the four with a direct molecular binding partner identified in a cell-free system, casein kinase 2 ^[6]. Its human evidence is a prospective multicentre cohort of 94 chronic haemodialysis patients, followed for a median of 26.5 months, in which circulating MOTS-c was independently associated with mortality and non-fatal cardiovascular events ^[7]. Nobody in that cohort was given anything. No human interventional study of MOTS-c has been published at all.

Tirzepatide has the full set. A 72-week placebo-controlled trial in 2539 adults with obesity ^[16], head-to-head randomised comparisons against semaglutide (Ozempic, Wegovy) in type 2 diabetes ^[17] and in obesity ^[13], and a nine-trial safety meta-analysis pooling 9871 participants ^[15]. The questions left open are downstream questions, and they remain design questions.

CJC-1295 stops at the surrogate. Its human work is real and was run in volunteers under supervision, but the endpoints measured were growth hormone and IGF-1 ^[21][22] — the intermediate, not the thing a reader cares about. No published trial asked whether that hormone change produced a difference anyone would notice.

What this planner plans

The name needs a boundary, so here it is, stated once and applied everywhere.

What gets planned on this site is a study. The unit is a trial specification: a population, a comparator, an endpoint, a duration, and enough participants to detect the effect being claimed. Every compound page closes with one, written against what the record already contains.

What does not get planned is anyone's use of anything. No quantities, no intervals, no combinations, no sequences intended for a person. Where a number appears in these pages it belongs to a published experiment and is reported as a description of that experiment — the amount someone was given under supervision so that a measurement could be taken. It is not carried forward as guidance, and the arithmetic that would convert it into guidance is not performed here.

The line is not cosmetic. Three of these four compounds have no approved human indication, which means there is no dosing standard to report even in principle. That absence is itself one of the findings, and it is the reason the missing study is worth specifying rather than lamenting.