How the register is built — and how it stays correct.
A peptide dosing reference is only useful if you can trust the math, the citations, and the cadence. This page documents all three.
Every compound carries a grade.
Each entry in the register is graded by the highest-quality evidence available to date, on a five-step ladder:
- FDA-approved
- An FDA-approved indication exists for this molecule. Off-label use is still possible, but the molecule itself has cleared regulatory review.
- Phase 3
- One or more randomised controlled trials at Phase 3 scale have published. Drug is post-Phase 3 or awaiting regulatory action.
- Phase 2
- Phase 2 RCTs have published. Evidence for efficacy exists; long-term safety and optimal dose may still be settling.
- Pre-clinical
- Animal or in-vitro data dominates. Human exposure data is limited to small open-label cohorts, case series, or clinic protocols.
- Research use only (RUO)
- No FDA-approved indication exists and human trial data is thin. Doses are derived from clinic protocols, vendor literature, or first-principles extrapolation. The catalog labels these explicitly.
Current distribution across 57 compounds in 6 classes — recomputed on every build.
Where the numbers come from.
The register applies a deterministic rule to every compound:
- If a published per-kilogram trial dose exists, the register scales by stored weight in kg.
- If only a fixed clinical dose has been published, the register shows the catalog default and labels it as such. It does not silently invent a scaled value.
- If a titration ladder has been published (semaglutide, tirzepatide, retatrutide), the register runs the published ladder week-by-week with dose holds and step-downs for tolerability.
- If a compound is sex-dimorphic (e.g. PT-141 / bremelanotide), the register applies the dose that matches the user's stored sex and labels the difference.
Vial reconstitution is also deterministic. Given vial mg and BAC water mL, the math computes mg/mL and volume per dose, then converts to U-100 syringe units rounded to the half-unit. The PDF export and the on-screen display always render from the same source.
What counts as a source.
Each catalog entry exposes a citations field. The priority order, highest to lowest:
- Published peer-reviewed trial (PMID, PMC, journal volume).
- Registered clinical trial (ClinicalTrials.gov NCT number).
- Manufacturer prescribing information for FDA-approved molecules.
- Named clinical protocol from a credentialed clinic (cited by name).
- Industry reference (Innerbody, vendor COA) — used only when nothing above is available, and only paired with a lower evidence grade.
Where a dose figure is contested across sources, the register selects the source with the strongest evidence grade and notes the alternative in the compound's notes field.
How the catalog stays current.
The register is reviewed against the literature on a weekly cadence. Each pass looks for:
- Newly published trials that change a compound's evidence grade.
- FDA action that promotes a molecule from RUO or pre-clinical.
- New safety signals or label updates that should change the monitoring panel.
- Errata in the catalog — reader-submitted corrections via support@44protocol.com are processed in the same pass.
Material changes ship to the public changelog at /changelog. Patch-level edits (typo, clarified phrasing) are batched.
What the register enforces on the user.
Three safety surfaces run continuously in the subscriber app:
- First-time-user half-dose notice. Users who self-identify as peptide-naive see a medical-voice notice recommending the first week at half the catalog dose, then titration up. The banner persists in the app and prints on every PDF export.
- Weekly safety review. A scheduled checklist surfaces every monitoring item and every serious adverse event across the user's current stack. The reminder triggers seven days after the last review and requires sign-off on every item to clear.
- Lab cross-reference. When the user adds a lab result, the register matches the marker text against the monitoring list of every compound in their stack and flags overlap.
How we prove the math is right.
Every build of the application runs a fuzz harness against the PDF and dose engines. The current harness asserts:
- Catalog parity across all 57 compounds with extreme weight and override inputs.
- Random-stack invariance across 200 procedurally generated stacks.
- Titration ladders for semaglutide, tirzepatide, and retatrutide week-by-week against their published schedules.
- Vial reconstitution math (mg/mL, mL per dose, U-100 units to the half-unit) across the entire catalog with custom vial-mg overrides.
The harness currently runs 7,000+ assertions per build. Any failure blocks deployment.
Who is responsible.
44 Protocol is operated by Stoneveil Holdings LLC, a registered Arizona company headquartered in Phoenix. Editorial oversight, software engineering, and customer support are run by the same small team.
For corrections, citations you think should be added, or questions about a specific dose, write to support@44protocol.com. We answer every legitimate inquiry within two business days.
What the register is not.
- Not medical advice. The register does not establish a clinician-patient relationship.
- Not a prescription system. We do not prescribe, dispense, source, ship, or sell any compound.
- Not a substitute for physician supervision. Every protocol must be reviewed with a licensed physician before initiation.
- Not a regulatory authority. Where a compound is unscheduled, scheduled, RUO-only, or otherwise restricted in your jurisdiction is your responsibility to determine.
If you find an error in the catalog, please write to us. We treat errata as P0.