Méthodologie v1.2.
Publiée le 2026-04-01. Renforce le poids des données humaines après Reta-1 et Surv-3, relève le plafond de fiabilité vendeur, et supprime le bonus WADA hérité qui comptait deux fois la posture réglementaire.
What changed
- 01
Human-data weighting raised 20% → 25%
Reta-1 NEJM publication and the Surv-3 readout pulled forward the case for weighting registered human work harder. Compounds with rodent-only files now lose a full point against compounds with even small Phase 1s.
- 02
Vendor-trust ceiling raised 7 → 9
Three independent third-party assays now report on the largest research-channel suppliers. The previous ceiling under-rewarded compounds with publicly verifiable lots.
- 03
WADA double-count removed
v1.1 captured WADA status both in regulatory posture and as a separate safety penalty. v1.2 consolidates that signal into regulatory posture only.
- 04
Mouse-only literature soft-cap
Evidence column is now capped at 8.0 when no published in-vitro-to-human bridge exists. BPC-157 and TB-500 both reset under this rule.
Weights, current
| Component | Weight v1.2 | v1.1 |
|---|---|---|
| Preuves publiées | 25% | — |
| Plausibilité du mécanisme | 15% | — |
| Données humaines | 25% | — |
| Fiabilité du vendeur | 15% | — |
| Statut réglementaire | 20% | — |