Species Dose Scaling Explainer
Convert a reported animal-model peptide dose to a Human Equivalent Dose (HED) using the FDA body-surface-area formula. The tool demonstrates the arithmetic — and prominently states why the answer is unreliable for peptides.
Educational only. Not a dose-recommendation calculator. Not for planning administration.
Animal dose
10 mg/kg
Mouse (20 g)
Human Equivalent Dose (HED)
0.811 mg/kg
= 48.65 mg for 60 kg adult
With safety factor ÷10
0.0811 mg/kg
= 4.865 mg for 60 kg adult
Calculation: HED = animal-mg/kg × (animal-Km / human-Km) = 10 × (3 / 37) = 0.811 mg/kg
Peptides are the class HED conversion works worst for.
FDA HED is a safety-margin heuristic for the first human dose of a small-molecule drug entering a Phase I trial — not a reliable pharmacological predictor. For peptides specifically, it fails badly because absorption, half-life, receptor-binding affinity and protein-binding all differ non-allometrically between species. Do not treat this output as a “research dose”. It is a teaching tool to demonstrate why rodent findings do not translate cleanly to human protocols. See our allometric-scaling-failures page for worked examples.
The FDA Km formula
The FDA’s 2005 guidance Estimating the Maximum Safe Starting Dose in Initial Clinical Trials for Therapeutics in Adult Healthy Volunteers normalises interspecies doses by body surface area rather than mass. Each species has a “Km factor” (roughly weight-in-kg ÷ surface-area-in-m²). The conversion is:
HED (mg/kg) = animal (mg/kg) × (animal Km / human Km)
For a mouse (Km 3) and human (Km 37): a 10 mg/kg mouse dose becomes 10 × 3/37 ≈ 0.81 mg/kg in human. Multiply by body weight to get a total dose.
Why the answer is misleading for peptides
The Km formula was derived from small-molecule toxicology data — drugs that distribute freely into aqueous compartments and clear through renal or hepatic pathways whose scaling roughly tracks body surface area. Peptides break this in several ways:
- Absorption from subcutaneous / intramuscular sites depends on lipophilicity, formulation, and injection-site perfusion — not on body surface area.
- Plasma half-life is dominated by proteolytic enzymes and receptor-mediated clearance, both of which vary non-allometrically between species.
- Fatty-acid conjugation (semaglutide, tirzepatide, cagrilintide) creates albumin-binding depot effects that make plasma PK a slow-release curve rather than a decay curve.
- Receptor affinity at the target frequently differs between the animal orthologue and the human receptor — the same molar concentration produces different biological effects.
A more thorough treatment lives on the allometric-scaling-failures page, with worked examples from the published peptide literature.
What this tool is not
- Not a starting-dose recommendation. Even the FDA framework multiplies HED by a safety factor of at least 10 — and that’s specifically for a Phase I first-in-human trial with monitoring, not a home research protocol.
- Not a substitute for the actual pharmacokinetic study. Where a peptide has been characterised in humans (the Enebo 2021 cagrilintide phase-1b is a good example), the trial data is always a better guide than allometric extrapolation.
- Not a signal to trust any body of preclinical-only evidence. Rodent dose-response curves are a starting point for a research programme, not a substitute for one.