Where these
numbers come from.

What we publish, and what it is

The compound library here is the same data that ships inside the BioHack Track app, published in full rather than summarised. Each entry carries a half-life, typical route, common dose range, cycle length, and notes on reconstitution, storage and cautions.

Those fields describe what is commonly reported. They are not a consensus, they are not a guideline, and for a great many of these compounds there is no guideline to be had — a substantial share are research chemicals with no approved human dose anywhere in the world. Where that is the case the page says so in its own words rather than leaving you to infer it.

The source hierarchy

When a compound has a citable primary source, it is one of these, in this order of authority:

• An FDA label or prescribing information, via accessdata.fda.gov or DailyMed. This is the strongest thing we can point at, and where one exists it is the authority on dosing — not us.
• A regulatory listing, such as the FDA’s list of bulk drug substances nominated for use in pharmacy compounding under section 503A. Read carefully: this is a record that a substance is not an approved medicine.
• Published literature and trial registries — PubMed, ClinicalTrials.gov — and class-level documents such as the WADA Prohibited List.

Sources are attached per compound and shown under Primary sources on that compound’s page. A compound with none shows nothing there. We would rather show an empty section than pad it.

Search links are not citations

Every compound page also carries a Look up the research block. Those are searches — they run a query against PubMed, Europe PMC and ClinicalTrials.gov and show you whatever is there today. They are labelled as searches on every page that has them, because presenting a search result as though a specific study endorsed a specific dose would be a lie told by implication.

Three regulatory states, decided by the page’s own sources

Each compound page opens with a Regulatory standing line. It is derived from that page’s own citations and nothing else:

• Approved medicine — an approved product exists and its label is linked below.
• Not an approved medicine — the compound appears on a compounding-evaluation list, which is an explicit record of non-approval.
• No approved product cited — the honest majority. We have no regulatory citation either way, and we do not guess.

This is deliberately a claim about the page rather than about the world, which makes it something you can actually check. It is decided by what a source says, never by the domain it sits on: several compounds cite fda.gov precisely because the FDA declined to approve them, and keying off the domain would print “FDA-approved” at the top of exactly the wrong pages.

What the build refuses to ship

Some of the checks below exist because the thing they catch already happened once:

• A source on the wrong compound fails the build. Every compound carrying sources must have its own name, or a documented alias, somewhere in those URLs or labels. A fuzzy name match once put naltrexone’s FDA label on the aspirin page — plausible-looking, completely wrong, on a page that quotes a dose.
• Every stated figure must match the app. Marker counts, compound counts and prices are read from the same data the app uses, and the build fails if a number written into the prose disagrees with it.
• Contrast and layout are checked, not eyeballed. Text that would render in the page’s own background colour, and images whose declared dimensions do not match the actual file, both fail the build.

How “last updated” is decided

The date on a page is the date its content last changed, computed from a hash of the page itself — not the date the site was last rebuilt. Rebuilding without changing anything does not move it. On a compound page the date tracks that compound’s own record, so it moves when the facts move and not when the footer does.

What we do not do

• No advertising, no sponsorship, no affiliate links. Nobody pays to appear here, and no link on this site earns us anything. The only outbound links are to the sources named on each page and to the two app stores.
• No advertising or behavioural tracking, and no third-party analytics beyond our host. There is no Google Analytics, no advertising pixel and no social embed. The site is served by Cloudflare, and Cloudflare injects its own privacy-preserving Web Analytics script on page loads, which tells us how many people visited and how fast pages rendered. It sets no cookie, does not fingerprint the browser and does not follow anyone between sites. Apart from that and the typeface, no external asset is loaded.
• No named clinician reviewer. Plenty of health sites print a doctor’s name and photograph at the top of a page as a trust signal. We do not, because a byline is not review, and one attached to 208 pages of research-chemical dosing would imply a level of sign-off nobody has actually given. The per-page regulatory standing is what we offer instead: it is narrower, and it is true.
• No user-submitted doses, ratings or protocols. Nothing on a compound page comes from a forum, a vendor or a comment.
• No vendor recommendations. We do not sell compounds, and we do not tell you where to buy them.

Corrections

If something here is wrong, we want the correction more than we want to be right. Email info@creativeimages.media — a specific page and, if you have one, a source is enough. Factual errors on a compound page are fixed at the data level, which means the fix lands in the app as well as on the site.

Who writes this

BioHack Track is made by a small independent team, and the same people write the app, the library and this site. We are not clinicians and do not present ourselves as any. What we can offer is that every figure is traceable to something you can open, and that the things we do not know are marked as unknown rather than smoothed over.

See also: the compound library, our help guide, the Terms of Use and the privacy policy.