Column
Underpowered: what n=8 supports, and what it does not
Eight rats per arm can detect an effect the size of a barn door. It cannot detect the modest, fragile effects that translation depends on.
Independent·Ann Arbor, MI·Est. 2021
Field Standard
Evidence before interpretation.
An independent editorial review of preclinical peptide research, animal models, laboratory methods, and the limits of translation.
Feature·Evidence Review
One lineage of evidence: the BPC-157 tendon literature, read closely
Dozens of rodent papers describe accelerated tendon healing in rats given a gastric pentadecapeptide. Nearly all of them trace back to a single research group in Zagreb. What does that concentration of authorship mean for the strength of the claim?
Latest Dispatches
Explainer
What a murine colitis model can and cannot tell us
DSS-induced colitis in C57BL/6 mice is a workhorse of inflammatory biology — and a poor mirror of any single human disease.
History
How one unreplicated paper becomes “established” online
A citation is a reference, not a confirmation. Tracing a 2004 rodent finding through two decades of repetition shows the difference.
Methods
Copper peptides and the distance between fibroblast culture and intact skin
A cultured human dermal fibroblast is a cell in a dish, not a dermis. The GHK-Cu literature illustrates how far apart those things are.
Editorial Standards
What this publication is
Field Standard reads the preclinical peptide literature so a general reader does not have to take it on faith. We summarize what rodent studies and cell-culture experiments actually report — the species, the model, the sample size, the laboratory — and we say plainly where the evidence stops. When a finding rests on eight rats in a single laboratory, that is what we write. When a claim has circulated online for a decade without independent replication, we write that too.
We hold a small set of working rules. We name the species. We state the gaps. We do not cite what we have not read. We publish no dosing, sourcing, or human-use information, because that is not what this literature can support and not what this publication is for. We accept no advertising, no sponsorship, and no affiliate arrangements. When we get something wrong, the correction is appended to the article and entered in a public ledger — not quietly deleted.
“A rat tendon is not a human tendon, and a healed rat is not a treated person. The distance between those two facts is where most of our work happens.”
From our editorial standards
How we read a paper
01
Sample size
Most preclinical peptide studies enroll single-digit animals per group. Small samples inflate effect sizes and make null results invisible, so we report n before we report conclusions.
02
Blinding & randomization
Whether investigators randomized animals and scored outcomes blind is one of the strongest predictors of whether a result survives. Many papers we review never mention either.
03
Independent replication
A finding repeated by the laboratory that produced it is not replication. We track who confirmed a result, who failed to, and who never tried.
04
Endpoint selection
Histology scores, behavioral tests, and molecular markers are chosen, not given. We note which endpoints a study measured, which it emphasized, and which may have been left in the notebook.