The finding nobody publishes.
An independent review of preclinical peptide research.
Everything covered here is animal-model and cell-culture research, summarised for general education. Not medical advice. Nothing on this site describes human use.
Evidence Review · Vol. 6, No. 2
Evidence Review
Three decades of rat tendon and ligament experiments, overwhelmingly from a single research lineage. That is not fraud, and nobody here is alleging it. It is a much more ordinary problem — and it is the reason the field has not moved.
Jonah AdeyemiResearch Editor · February 18, 2026 · 7 min read
Eight entries · Last reviewed Feb 2026
Structured summaries of what the preclinical literature reports for each compound, and of where that literature stops.
No advertising · No sponsorship
The Null Result reads published preclinical papers about a small set of research peptides and reports what they actually found — in which species, in which injury model, at what scale. Almost all of that work is in rats, mice, and cultured cells. A result in a rat is a fact about rats. It generates a hypothesis about humans; it does not settle one, and we do not write as though it does.
We publish nothing about dosing, sourcing, preparation, or use in people, and we are not a buying guide, a forum, or a supplier. Our standards are unglamorous and we hold to them: we name the species and the model every time, we say plainly when a finding rests on a single laboratory or a handful of animals, we distinguish what has been replicated from what has only been repeated, and when a compound has essentially no rigorous human evidence we write that sentence rather than a softer one. Corrections run in public and stay attached to the piece. More about who writes this →
A mouse that heals is a fact about a mouse. Everything after that sentence is inference — and inference is where this field keeps overspending. The Null Result editorial desk
Four questions, in order
We look for how many animals were in each arm before we look at what happened to them. Small groups do not merely produce uncertain estimates — when they do reach significance, they systematically overstate how large the effect is.
Much of the peptide literature scores injury or healing by eye, and much of it does not report whether the scorer knew which animal was which. Where allocation and assessment are not described as blinded, we say so rather than assuming it was done and left out.
Repetition inside one laboratory is not replication; it shares the model, the reagents, the surgical hand, and the assumptions. We track whether an unaffiliated group in a different institution has run the experiment and reported the outcome either way.
A study measuring tensile load at failure and a study scoring a histology slide are not measuring the same thing, even when both report improvement. Heterogeneous endpoints across a literature make it look more consistent than it is, because nothing can contradict anything else.