The DaX™ platform, pioneered by Caltech Professor David Van Valen, MD, PhD, is a first-in-class, structure-based protein design model for Mirror Peptides. The platform harnesses advancements in Generative AI to unlock the insights of natural ligand-receptor data sets to build de novo Mirror Peptide structures that bind to clinically-relevant targets. DaX™ leverages the latest methods in guided diffusion, computational binding models, and GPU compute architecture to advance this novel class of medicine.

With demonstrated success against multiple transmembrane receptor targets, DaX™ delivers remarkable drug development efficiency: design-build-test cycles in ~4 weeks with laboratory validated hit rates higher than current industry standards. The platform's sophisticated multi-objective optimization enables simultaneous refinement of key drug properties, including specificity, half-life, and solubility, transforming the scope and velocity of drug development.