Project 01 · Digital Skin
247,861 molecules tested on digital skin — each one replacing a ≈€5,000 lab study. A model that synthesizes human skin and predicts how compounds behave in it, in silico.
Proof
0.84
R² against held-out measurements.
2,149
real Franz cell measurements behind it.
10,000×
cheaper than a single ≈€5,000 wet-lab study.
<48h
from candidate to filed patent.
Live simulation
Validated blind
Sixty-seven held-out compounds across six chemical classes, one model. Predicted permeation plotted against measured permeation lands on the diagonal — R² 0.84 against numbers it never trained on. Hover any point for its values.
How it compounds
10,000 virtual skin tests overnight, each replacing a €5,000 Franz cell study. The pipeline generates candidates, searches patents, simulates permeation, and scores safety and viability.
Composition-of-matter claims covering the molecule, its series, and its applications — portfolios that generate royalties for decades.
Partners synthesize, validate, and sell; we collect on every kilogram. One successful ingredient: €200K–5M a year.
First discovery
A prodrug of niacinamide — the $2.5B skincare active — engineered for 307× improved skin penetration. It releases niacinamide directly in the epidermis via skin amidases (t½ 5.2h). Discovered computationally, validated by expert chemists, entering synthesis.
307×
Permeation · vs niacinamide (validated)
26.5%
Epidermis delivery
Amidase
Prodrug, activated in skin
Clean
Low irritation · no phototox
Niacinamide · the $2.5B active
−8.94 logKp
Very low permeation
N-Octyl Nicotinamide · ours
−6.44 logKp
307× faster
Built on HuskinDB · SkinPiX · EDETOX · 2,149 real Franz cell measurements · OECD TG 428 compliant
Work with us
We’re looking for manufacturing partners — cosmetic brands, ingredient suppliers, and CDMOs who can synthesize, validate, and bring novel ingredients to market. We provide the discovery, the patent, and the science. You provide the lab and the shelf. Royalties are shared.
Request access →