Methodology v1.1
The rubric Pepvise used through Q1 2026, superseded by v1.2 on 2026-04-01. Maintained as a historical reference. No review in the current database is scored against v1.1; this page is here so the methodology pill on any archived or syndicated v1.1 review keeps resolving.
For the current rubric, see methodology v1.2.
The five dimensions (v1.1)
- 01
Evidence (30%)
Quality and quantity of preclinical, in-vitro, and animal-model evidence. Capped at 9.0 in v1.1; v1.2 introduced the 8.0 soft-cap for compounds without an in-vitro-to-human bridge.
- 02
Mechanism (20%)
Whether the proposed mechanism of action is a coherent extension of established cell biology or a marketing back-fit. Unchanged between v1.1 and v1.2.
- 03
Human data (20%)
Published human work: pilot studies, Phase 1 dose-escalation, Phase 2 dose-finding, Phase 3 efficacy. v1.2 raised the weighting to 25% to better reflect the gap between rodent files and registered human work.
- 04
Vendor trust (20%)
Whether a quality-controlled commercial supply path exists. Ceiling 7.0 in v1.1; raised to 9.0 in v1.2 after three independent third-party assays began publishing on the largest research-channel suppliers.
- 05
Safety + regulatory posture (10%)
FDA / EMA approval status, WADA listing, boxed warnings, withdrawal history. v1.1 double-counted WADA status here AND as a separate safety penalty; v1.2 consolidates that signal into regulatory posture only.
What changed in v1.2
- 01
Human-data weighting raised 20% → 25%
The 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
The previous ceiling under-rewarded compounds with publicly verifiable lots. Three independent third-party assays now report on the largest research-channel suppliers.
- 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.
Re-scoring policy
When a methodology version bumps, existing reviews are not auto-re-scored. The pill on each review records the version it was scored under at the time of last editorial review. A review re-scores against the current rubric only when its lastUpdated field is bumped and the editor explicitly applies the new dimension weights. This protects the audit trail: every score on every review traces to a specific methodology version, a specific date, and the editor who signed it off.