Corrections

When we get something wrong and it is pointed out, the fix and the report are both logged here. If the person who reported it has a stake in the compound or the paper, that is recorded too. How we classify research is on How we rate.

Catalogue-wide: family absorption
Reported Sep 4, 2026 · Fixed Sep 4, 2026
What was wrong
Pages were absorbing the literature of their molecular family. A single passing mention of a compound was enough to link a paper, so receptor-family reviews, doping-assay method papers and sibling-compound studies were being counted as evidence for the compound itself. Four pages had already been caught this way one at a time (Ipamorelin/ghrelin, P21/CDKN1A, PT-141/melanocortins, NAD+/precursors).
What changed
Ran one systematic pass over all 85 published pages: a linked paper is treated as absorbed when its title never names the compound and the whole text names it at most once. 598 such links were removed across 77 pages (largest: Cartalax 35, CJC-1295 20, Teriparatide 18, Somatropin 17, Cagrilintide 16, Ipamorelin 15, Sermorelin 13, Vesugen 13). Study records were not deleted, only unlinked. Matching was also made separator-insensitive so TP508/TP-508 and NAD/NAD+ are the same term, and synonyms were added for Somatropin, SNAP-8, 5-Amino-1MQ and Thymosin Beta-4. Five genuine human trials caught by the heuristic were restored and locked (three GHRH cognition trials on Tesamorelin, two Thymosin Beta-4 rows), and the Thymosin Beta-4 seborrheic dermatitis trial was re-locked as human after a reclassification run had reset it. Editor-locked rows are now exempt from both the pass and the on-topic check.
Reported by
Site owner — PeptideSearch
NAD+
Reported Sep 4, 2026 · Fixed Sep 4, 2026
What was wrong
The NAD+ page held the research on two different molecules. Of 27 rows labelled studies in people, only three gave anyone NAD+ itself: ten gave nicotinamide riboside, ten gave nicotinamide mononucleotide, one compared boosters head to head, and four gave nothing at all. One row labelled a study in people was done in rats.
What changed
Nicotinamide riboside and nicotinamide mononucleotide now have their own pages, and every row moved to the compound that was actually given. The four rows where nothing was administered were removed, and the rat study was relabelled and locked.
Reported by
Dre Smith
NAD+
Reported Sep 4, 2026 · Fixed Sep 4, 2026
What was wrong
The NAD+ page counted 27 human studies. Most of those studies gave people a different molecule: nicotinamide riboside, nicotinamide mononucleotide, or plain nicotinamide. Four gave people nothing at all. Only four studies actually gave people NAD+ itself.
What changed
Split the library into separate pages for nicotinamide riboside and nicotinamide mononucleotide, and moved those studies there. Removed the four studies that gave nobody anything, and removed one study of plain nicotinamide (vitamin B3), which is a different molecule again and did not warrant its own page on a single study. NAD+ now shows 4 human studies. The head-to-head comparison of all three is listed on all three pages and is flagged so it is not counted three times.
Reported by
Dre Smith — PeptideSearch editorial
GHK-Cu
Reported Aug 26, 2026 · Fixed Aug 26, 2026
What was wrong
The 503A status was shown as a flat Category 1. The FDA entry reads "GHK-Cu (except for injectable routes of administration)", so a flat Category 1 implied injectable GHK-Cu was covered when it is not.
What changed
The status is now stored as a route-qualified entry and rendered as Category 1 for all routes except injection, with the verbatim FDA entry quoted.
Reported by
Dre Smith — PeptideSearch
GHRP-2
Reported Aug 26, 2026 · Fixed Aug 26, 2026
What was wrong
The 503B status was shown as a flat Category 2. FDA splits the entry: Category 1 for all routes except injectable and nasal, Category 2 for those two routes. A flat Category 2 said every route was restricted when most are not.
What changed
Both 503B entries are now stored separately and rendered as two lines, each with its routes and the verbatim FDA entry.
Reported by
Dre Smith — PeptideSearch
PT-141 (bremelanotide)
Reported Aug 26, 2026 · Fixed Aug 26, 2026
What was wrong
Five papers on the page were about the melanocortin receptor family or other melanocortin compounds rather than bremelanotide: an NDP-MSH mouse nephropathy study tagged as an animal study, an in-silico SARS-CoV-2 docking screen and a melanocortin-receptor ligand review tagged as reviews, a macrocycle medicinal-chemistry paper, and a "mercy sex" analysis of pooled HSDD trial baseline data.
What changed
The five rows were removed from the page. The MC4R cryo-EM structure paper, which does resolve bremelanotide bound to the receptor, was kept and reclassified from unclassified to a lab-dish (structural) study. Human study count is unchanged at 11.
Reported by
Dre Smith — PeptideSearch
GHRP-2
Reported Aug 24, 2026 · Fixed Aug 25, 2026
What was wrong
We stated that the FDA had kept GHRP-2 in its restricted category and that it was not part of the July 2026 review. That is wrong for 503A: FDA has "GHRP-2" in 503A Category 3 — nominated without adequate support — per the FDA 503A category document dated May 14, 2026. Restricted was the wrong category, and the review claim had nothing behind it.
What changed
The page now shows FDA Category 3 for 503A compounding, which means the evidence sent to the FDA was not enough to judge it — a different thing from a safety restriction. The unsupported claim about the July 2026 review has been removed. The FDA history is recorded on the page timeline with the source document.
Reported by
Internal review — None. Found by our own audit of the FDA category documents.
GHRP-6
Reported Aug 24, 2026 · Fixed Aug 25, 2026
What was wrong
We stated that the FDA had kept GHRP-6 in its restricted category and that it was not part of the July 2026 review. FDA has "GHRP-6" in 503A Category 3 — nominated without adequate support — per the FDA 503A category document dated May 14, 2026.
What changed
The page now shows FDA Category 3 for 503A compounding and no longer claims a safety restriction or the July 2026 review. The FDA history is recorded on the page timeline with the source document.
Reported by
Internal review — None. Found by our own audit of the FDA category documents.
MGF
Reported Aug 24, 2026 · Fixed Aug 25, 2026
What was wrong
We stated that MGF is not on the U.S. compounding list. FDA has it as "Mechano growth factor (MGF)" in 503A Category 3 — nominated without adequate support — per the FDA 503A category document dated May 14, 2026.
What changed
The page now shows FDA Category 3 for 503A compounding and names the FDA entry. It still says a doctor cannot prescribe it, which Category 3 supports. The FDA history is recorded on the page timeline with the source document.
Reported by
Internal review — None. Found by our own audit of the FDA category documents.
Thymosin Alpha-1
Reported Aug 24, 2026 · Fixed Aug 25, 2026
What was wrong
We stated that Thymosin Alpha-1 is restricted from 503A compounding. There is no basis for that. It does not appear anywhere in the FDA 503A category document dated May 14, 2026 — we checked every page of the document, including across page breaks, and there is no entry under Thymosin, Thymalfasin or any alpha-1 form. We asserted a restriction that FDA never made.
What changed
The claim has been removed. The page now says nobody has put it forward for 503A compounding, so there is no 503A pathway and no 503A restriction either. Its only appearance is on the separate 503B list for large outsourcing facilities, recorded on the page timeline.
Reported by
Internal review — None. Found by our own audit of the FDA category documents.
GHK-Cu
Reported Aug 24, 2026 · Fixed Aug 25, 2026
What was wrong
Our wording said GHK-Cu is "on the 503A bulks list". The facts we published were right but the phrasing was not: Category 1 is a list of substances that have been put forward and are still being evaluated, not the final bulks list, which is set by rulemaking.
What changed
Category 1 is now described site-wide as still under FDA evaluation, with the FDA position that it does not intend to act against a pharmacy compounding it while the review runs. No factual claim about GHK-Cu changed. The published video already used this wording.
Reported by
Internal review — None. Found by our own review of site wording.
MK-677
Reported Aug 24, 2026 · Fixed Aug 25, 2026
What was wrong
We stated that MK-677 is not on the compounding list. FDA has it under its chemical name "Ibutamoren Mesylate" in 503A Category 2 — the group it says raises significant safety risks — per the FDA 503A category document dated May 14, 2026. Our check missed it because we were not matching the chemical name.
What changed
The page now shows FDA Category 2 for 503A compounding and names Ibutamoren Mesylate as the FDA entry. The FDA history is recorded on the page timeline with the source document.
Reported by
Internal review — None. Found by our own audit of the FDA category documents.
ARA-290
Reported Aug 24, 2026 · Fixed Aug 25, 2026
What was wrong
We stated that ARA-290 is not on the compounding list. FDA has it as "Cibinetide (ARA-290)" in 503A Category 3 — nominated without adequate support — per the FDA 503A category document dated May 14, 2026.
What changed
The page now shows FDA Category 3 for 503A compounding and names the FDA entry. The FDA history is recorded on the page timeline with the source document.
Reported by
Internal review — None. Found by our own audit of the FDA category documents.
GHK-Cu
Reported Aug 24, 2026 · Fixed Aug 24, 2026
What was wrong
We stated that GHK-Cu is not on the list a pharmacy can compound from. That is wrong. It is on the FDA 503A Category 1 list for all routes of administration except injectable, per the FDA 503A Bulk Drug Substances Category 1 list, document dated May 14, 2026. The error appeared on the GHK-Cu peptide page and in the published video.
What changed
The page now says a compounding pharmacy can legally prepare GHK-Cu for topical use but not for injection, and carries the FDA history: removed from Category 1 on April 22, 2026 when the nominations were withdrawn, then added back on May 5, 2026 after one nominator clarified it withdrew only the injectable-route nomination. FDA intends to consult the PCAC before the end of February 2027. The video was re-recorded and republished.
Reported by
Internal review — None. This was found by us, not reported from outside.
NAD+
Reported Aug 24, 2026 · Fixed Aug 25, 2026
What was wrong
We stated that no doctor can prescribe NAD+ as an approved medicine and that it is not on the compounding list. The second half is wrong. FDA has "Nicotinamide Adenine Dinucleotide (NAD)" in 503A Category 1 — substances under evaluation — per the FDA 503A category document dated May 14, 2026. This is the second error found on this page.
What changed
The page now says NAD+ is in FDA Category 1 for 503A compounding: still under evaluation, with FDA stating it does not intend to act against a pharmacy that compounds it while the review runs. The page no longer implies there is no compounding pathway, and it makes clear that Category 1 is not the final bulks list. The FDA history is recorded on the page timeline with the source document.
Reported by
Internal review — None. This was found by our own audit of the FDA category documents, not reported from outside.
Glutathione
Reported Aug 24, 2026 · Fixed Aug 25, 2026
What was wrong
We stated that an FDA advisory panel voted in favor in 2022 and that it is still not on the list. The second half is wrong. FDA has "Glutathione" in 503A Category 1 — substances under evaluation — per the FDA 503A category document dated May 14, 2026.
What changed
The page now shows FDA Category 1 for 503A compounding, with the point that Category 1 is not the final list and the rulemaking that would put it there is still unfinished. The FDA history is recorded on the page timeline with the source document.
Reported by
Internal review — None. Found by our own audit of the FDA category documents.
VIP
Reported Aug 24, 2026 · Fixed Aug 25, 2026
What was wrong
We stated that VIP is not approved and said nothing about compounding, which read as no pathway existing. FDA has "Vasoactive Intestinal Peptide" in 503A Category 1 — substances under evaluation — per the FDA 503A category document dated May 14, 2026.
What changed
The page now shows FDA Category 1 for 503A compounding, with the FDA position that it does not intend to act against a pharmacy compounding it while the review runs, and that this is not the final list. The FDA history is recorded on the page timeline with the source document.
Reported by
Internal review — None. Found by our own audit of the FDA category documents.
Kisspeptin-10
Reported Aug 24, 2026 · Fixed Aug 25, 2026
What was wrong
We stated that the FDA advisory panel reviewed it in October 2024 and did not recommend it for the compounding list. That understates the position. FDA has "Kisspeptin-10" in 503A Category 2 — the group it says raises significant safety risks — per the FDA 503A category document dated May 14, 2026.
What changed
The page now shows FDA Category 2 for 503A compounding and states plainly that pharmacies are not allowed to make it. The FDA history is recorded on the page timeline with the source document.
Reported by
Internal review — None. Found by our own audit of the FDA category documents.
Multiple peptides
Reported Aug 23, 2026 · Fixed Aug 23, 2026
What was wrong
A study testing the purity and contents of peptides sold directly to consumers appeared unclassified on several peptide pages, including GHK-Cu, CJC-1295, Tesamorelin, PT-141 and others.
What changed
Added a category for product testing: "Testing of products for sale, not a health study". It is now applied consistently everywhere that paper appears, and ranks alongside lab-method papers rather than as evidence about health.
Reported by
Site reader — —
Tesamorelin
Reported Aug 23, 2026 · Fixed Aug 23, 2026
What was wrong
The 2007 New England Journal of Medicine randomized, placebo-controlled trial that Egrifta was approved on was labelled "Data from people, not a trial". Its randomization is described only in the abstract, not the title, so our classifier missed it. The 2008 AIDS safety extension study was left unlabelled for the same reason.
What changed
Both are now labelled as studies in people. The rule that assigns "Data from people, not a trial" now requires positive evidence of an observational design, and trial evidence is read from the abstract (random assignment, trial registration numbers) rather than the title alone. A sweep of every affected record reclassified 70 papers across the site.
Reported by
Site reader — —
NAD+
Reported Aug 22, 2026 · Fixed Aug 22, 2026
What was wrong
The study "Oral LNAD+ rapidly elevates whole blood intracellular NAD and metabolic flux without elevating plasma NAD: evidence from a randomized controlled trial" (GeroScience, 2026) was labelled "Data from people, not a trial". It is a double-blind, randomized, placebo-controlled Phase 0/1b trial (NCT07336836, 60 randomized, primary analysis n=50, healthy adults 45-75).
What changed
The row is now labelled "Study in people". The cause was our classifier reading "retrospectively registered" in the abstract as a retrospective study design, which overrode "randomized controlled trial" in the title. A trial design stated in the title now settles the classification, and two further rows were corrected the same way.
Reported by
Michael Leitz-Langan MD — Bryleos; listed as an author on the paper in question.