• Automated Reaction Simulations: Our computational scientists ran high throughput workflows to model DNMT1’s reaction dynamics and pinpoint its transition state*.

• Pharmacophore Extraction: Using our proprietary Quantum Pharmacophore technology, we captured the shape, charge distribution, and interaction features that uniquely characterize DNMT1’s transition state — setting it apart from closely related enzymes.

• Generative AI Molecule Design: These quantum-derived features were fed into our generative AI engine, which proposed novel molecules tailored to bind the transition state with high specificity.

• Expert-Informed Prioritization: Combining computational scoring, structural visualization, and domain expertise, we selected top candidates for synthesis and experimental validation in the wet lab.

*We used so-called QMMM simulations which combine a quantum region to describe the chemistry and a less accurate but cheaper classical model for the wider enzyme.