About the publication
Founded in Ann Arbor in 2021 by two people who kept finding that the only readable explanations of preclinical peptide research came from businesses selling something.
Mara Ellison spent nine years as a laboratory animal technician at a university vivarium, running behavioural assays on other people’s grants. She was the person who weighed the animals, scored the tests, and typed the numbers into somebody else’s spreadsheet, which meant she was also the person who noticed when a paper came out and described those numbers as considerably more than they had been. Clara Vasquez had spent the preceding decade as a science journalist covering supplement and nutrition claims, where the pattern was familiar: a real study, a modest finding, and an enormous industry built on the gap in between.
They met in 2021 through a mutual acquaintance who thought they would find each other exhausting. Instead they discovered they had independently been keeping the same kind of file — a running document of claims about research peptides, alongside notes on what the underlying rodent papers had actually measured. Clara’s version was better organised. Mara’s had more marginalia, most of it profane.
For most of that first year they worked at the same table at Argus Farm Stop on Fourth Avenue, Tuesday and Thursday mornings, because it had reliable power outlets and nobody minded two people arguing quietly about histology scoring for three hours. The shared document grew until it stopped being a document. In November 2021 they published the first four entries from it — BPC-157, thymosin β4, GHK-Cu, and KPV — as the beginning of what is now the Library. The site had no logo for another six months.
In the second year the sponsorship offers started. Three arrived in one month, all from companies selling the compounds we were writing about, all polite, all substantial relative to what this operation costs to run. We turned them down and then wrote a policy so we would not have to relitigate it every time: no advertising, no sponsored content, no affiliate arrangements, no paid placement of any kind, ever. It has cost us money and saved us the only thing we actually have.
In March 2023 we ran a correction on a piece about rodent cardiac work that had described results without noting they were produced using a modified, more potent analogue rather than the native peptide. The correction was appended to the article with the original wording left visible, as ours always are. The contributor who wrote it felt the public appendix was disproportionate to the error and stopped writing for us. We think he was wrong, we understand why he left, and we would run it the same way again. The alternative — quietly editing the sentence and moving on — is precisely the habit this publication exists to argue against.
The name is a small methodological joke that turned out to be the mission statement. In any experiment, the control group is the arm you don’t treat: the baseline that makes the treated arm mean anything at all.
Mission
We summarise and criticise the published preclinical literature on a small set of peptides, naming the species and the model behind every result. We state plainly where the evidence is thin, unreplicated, or absent, and we sell nothing to anyone.
Scope & limitations
We cover preclinical, animal-model research: studies in rats and mice, work in cultured cells, and the methodology underlying both. Eight peptides are in scope, listed in the Library. Everything we publish about them is a description of laboratory research, written for a general reader who wants to know what the papers say and how much weight they will bear.
Several things are deliberately excluded, and the reasons are worth stating rather than leaving as a gap:
We publish no quantities, schedules, routes, cycles, or preparation instructions of any kind. Animal studies use quantities scaled to animals under laboratory conditions, and reformatting them into something that reads like instructions would misrepresent both the research and our confidence in it.
We do not name vendors, compare products, discuss testing or purity, or link anywhere a substance can be obtained. We are a publication, not a directory, and the moment we became either we would lose the only thing that makes our assessments worth reading.
We do not describe, evaluate, or collect accounts of people taking these compounds. Self-reports are not data, and treating them as though they were is how this subject became unreadable in the first place.
Licensed metabolic medicines and the clinical pharmacology around them are outside our remit entirely. Those fields have enormous registered human trial literatures and are well covered by the clinical and pharmacological press; we cover thinly evidenced preclinical work, which is a different problem requiring a different kind of scepticism.
Nothing we publish states or implies that any compound benefits a person. Rodent results are hypothesis-generating. The distance between a mouse and a patient is the subject of this publication, not a technicality to be waved through.
Editorial standards
Every reported result carries its model: the strain, the species, the injury or disease preparation, and whether it happened in an animal or a dish. “Improved healing” is not a finding until you know it was a histological score in a rat with a surgically transected tendon.
Where a compound has no controlled human evidence, we write that sentence explicitly rather than ending the summary and letting the silence do the work. Absence of evidence is information, and it is usually the most important information on the page.
No identifier appears in our pages for a paper nobody here has read in full. When we describe a literature at the level of models and dates rather than individual references, that is a deliberate signal about the depth of our reading, not an oversight.
No quantities, no schedules, no suppliers, no purity comparisons, no links to anywhere selling anything. This rule has cost us traffic and it is not negotiable.
Corrections are appended to the article, dated, with the original wording left intact and visible. We do not silently edit. A publication that quietly fixes its record has no record.
No display advertising, no sponsored posts, no affiliate links, no paid placement, no gifts from anyone with a commercial interest in these compounds. Contributors declare any relevant relationship before assignment, and we decline assignments where one exists.
The masthead
Nobody here holds a clinical credential, and nobody here is pretending to. What the desk has instead is time spent close to how this research is actually produced.
Nine years as a laboratory animal technician in a university vivarium, running behavioural assays for investigators whose papers she later read with growing discomfort. She knows what a Tuesday in an animal facility looks like, which is the single most useful qualification for reading this literature and the rarest among people who write about it. She edits every Library entry and argues with every writer about the verb tense.
Left a neuroscience PhD after the central finding of his thesis dissolved once the scoring was properly blinded — an experience he describes as the most valuable eighteen months of his education and the worst year of his life. He writes almost exclusively about how much work the word “promising” is doing in a given sentence, and about the gap between a result that has been repeated and one that has only been cited.
A decade covering supplement and nutrition claims taught her the shape of an evidence gap: a real study, a modest finding, and a marketing apparatus in the space between. She came to peptides because the shape was identical and the gap was wider. She writes the methods pieces, and is responsible for the house insistence that a model be described before its results are.
A working biostatistician who writes the recurring Power Analysis column about sample sizes and what they will and will not support. He is unfailingly polite about other people’s statistics and completely immovable on the arithmetic. His standing position is that most null results in this field are not evidence of absence but evidence that the study was never capable of finding anything.
A former medical librarian, which means she is the only person here who can reconstruct a citation chain backwards without complaining about it. She covers retractions, corrections, and the specific mechanism by which one weak paper becomes an established fact: the point in the chain where a hedge quietly falls off and never gets restored.
Twenty-two years on newspaper copy desks before this, and the reason nothing reaches the site with “proven,” “breakthrough,” or “miracle” in it. He checks every species claim, every date, and every characterisation of a study design against the writer’s notes, and he has a laminated list of verbs that require a source before they are permitted in a headline.
How we’re funded
The newsletter is free and the site has no paywall. Running costs are genuinely small — hosting, a domain, one privacy-respecting analytics subscription, and modest per-piece fees for contributing writers — and they are covered by voluntary contributions from readers plus the founders’ own money when contributions fall short, which they regularly do.
There is no advertising on this site and there will not be. We accept no sponsorship, no sponsored content, no affiliate commissions, no product samples, and no payment of any kind from any company that manufactures, distributes, or sells research compounds or anything adjacent to them. If that ever changes it will be announced on this page before it appears anywhere else, and the archive will be marked.
Corrections
Corrections are appended to the bottom of the affected article, dated, describing what was wrong and what it has been changed to. The original wording stays visible. We do not silently edit published work, and we do not remove pieces that have aged badly — we annotate them.
To request a correction, write to corrections@the-control-group.org with the article, the specific passage, and what you believe the accurate statement to be. Sourcing helps and is not required; we will do the checking either way. We respond to every correction request, including the ones we decline, and we say why.
If you want to write for us, pitches@the-control-group.org is the address. Send two paragraphs on the argument and what evidence it rests on, rather than a finished draft. We commission from outside the masthead regularly and pay a flat per-piece rate. We do not accept pitches from anyone with a commercial interest in the compounds we cover, and we ask directly.
Contact
General — hello@the-control-group.org
Editorial — editors@the-control-group.org
Corrections — corrections@the-control-group.org
Pitches — pitches@the-control-group.org
Corrections are triaged within two working days. General and editorial mail usually takes under a week. Pitches take longer, sometimes three weeks, because they are read properly by more than one person. We cannot answer questions about using any compound, and we do not make exceptions.