No Single Metric Tells the Whole Story
A shift in one marker can't tell you whether a molecule works — cell identity, function, and stress all matter too.
Skin Anti-Aging Target Discovery
A skin-specific virtual cell simulates molecular interventions to shortlist candidate molecules and multi-stage sequential regimens. Wet-lab validation and active learning keep refining the results.
The Challenge
The combinations explode
Molecules, doses, durations, and intervention sequences interact, making one-by-one testing impractical
Predictions Require Experimental Validation
Simulation results depend on data and model conditions; actual cellular responses still require wet-lab validation
Core Architecture
A skin-specific virtual cell that connects data, prediction, and experimental validation
R&D Deliverables
Experimentally validated candidate molecules and regimens
01
Candidate Molecules
Molecule list, screening priorities, and validation results, with the experimental system and applicable conditions noted.
02
Multi-Stage Sequential Regimens
Molecule combinations and sequencing, with the doses and durations used in validation and the measured cellular responses.
03
Prediction and Validation Evidence
Linked state predictions, evaluation metrics, and experimental results, with model versions and iteration rationale retained.
Core Value
Narrow the Candidate Pool
Improve Cellular Context Relevance
Strengthen the Mechanistic Evidence Chain
Optimize Validation Experiment Design
