This page states the rules PeptiDex writes to. They are written down so a reader can hold the site to them, and so a failure is a broken rule rather than a matter of opinion.
1. Null, not guessed
If a value is not established, the entry states nothing. It does not estimate, round, average, or infer.
This is the rule PeptiDex is most willing to be judged on, because it is the one that costs something. A blank row looks worse than a number. Filling it would make every entry look more complete, more authoritative, and more like the competition. It would also mean a reader could not tell which figures are real.
So the blanks stay, and here is the current count:
| Field | Entries stating a value | Entries stating nothing |
|---|---|---|
| Half-life | 43 | 36 |
| Molecular weight | 60 | 19 |
| Amino-acid sequence | 47 | 32 |
| Chain length | 49 | 30 |
Out of 79 entries. Those figures are generated from the data at build time, not typed into this page, so they cannot drift from what the library actually contains. Open any entry with a missing row and you will find the row simply absent from the specification table — not a dash, not "approximately", not "varies".
The same rule governs the pages' claims about themselves. A compound's spec table drops the "typical form" row entirely where the entry's own description contradicts the category default, rather than asserting a dosage form the text does not support.
2. Sources, and the absence of them
157 primary references are published across 68 compounds. Each one was retrieved from its PubMed record and checked on 2026-07-28: the identifier resolves, and the title, journal and year on our page match the record.
No citation is ever written from memory or reconstructed. A fabricated reference on a health-adjacent site is worse than no reference at all, because it converts an honest gap into a false claim that looks like diligence.
Where a compound has no indexed literature, the page says so:
No indexed primary literature located for this compound as of 2026-07-28.
That statement currently appears on 9 entries. It is not an oversight and it is not a placeholder — it is the finding. Several of those entries are supplier blends and compounded solutions that exist as products rather than as objects of study, and a reference library should be willing to say so.
Compounds with no reference block yet have not been reviewed for sources. Absence of a block makes no claim in either direction.
We also exclude what PubMed flags as retracted. At least one paper widely cited around one of these compounds has been retracted; it does not appear here, and a retracted paper will never be added to fill a gap.
3. Everything here is literature summary, not a finding
PeptiDex reports what has been published. It does not conduct research, run assays, test material, or verify supplier claims independently.
Most of the literature summarised on this site is preclinical — in-vitro work or animal models. A mechanism described in a cell line is not a mechanism established in humans. An effect reported in rodents is not an effect you should expect. Every entry carries that framing, and it is not boilerplate: for a large share of these compounds there is no human evidence at all.
Consequently:
- No therapeutic claims. Listing a research area is not a claim of efficacy and not an indication.
- No dosing guidance. Where amounts appear, they are amounts reported in published research and research-community protocols, given as educational context. They are not instructions, not recommendations, and not scaled for human use.
- No medical advice. PeptiDex is written for research and educational purposes only. See the disclaimer.
4. Independence and money
PeptiDex earns affiliate commission on purchases made through its outbound supplier links. That is disclosed on the about page, on the disclaimer, and next to the links themselves.
No supplier writes, reviews, or approves an entry before it is published, and no entry exists because a supplier asked for it: 33 of 79 entries carry no supplier link at all. A compound's presence in this library, the values in its table, and whether it gets a reference block are decided the same way whether or not there is anything to sell against it.
5. Review cadence
Every one of the 79 entries carries a dated review stamp. That date is
not decorative — it drives the page's dateModified and its lastmod in
sitemap.xml, so it is visible to anyone who wants to check it, and it is
deliberately not reset to the build date on every deploy.
The rules:
- An entry's review date is bumped in the same change that edits its facts. A change that quietly leaves the old date would be telling crawlers and readers the page had not changed on precisely the release that rewrote it.
- The full library is swept at least twice a year regardless of whether anything prompted it.
- A reported error is treated as its own trigger and is not held for a scheduled sweep.
The most recent review stamp in the library is 2026-07-28.
6. Corrections
Send corrections to contact@thepeptidex.net with the page and the specific claim you are disputing. A source helps and is not required — "this number looks wrong" is a perfectly good report.
What happens next:
- Factual errors are fixed and the entry's review date is bumped, which is what makes the correction visible from outside.
- Where a value turns out not to be established, it is removed rather than replaced. A corrected guess is still a guess. The blank is the correct answer.
- If we cannot establish who is right, the claim comes down until it can be sourced. A disputed statement does not get to sit on the page while it is argued about.
7. What this policy does not promise
It does not promise the site is complete, or current with a field that moves faster than any reference can. It does not promise a value stated here is correct — only that it was taken from a source rather than invented, and that we will remove it when shown otherwise.
If you find a page that breaks one of these rules, that is a bug. Please report it.