CIPHERICA Launches CIPHER-SKI: A New Paradigm in AI-Driven Drug Discovery
Portland, OR – May 19, 2026 — CIPHERICA today officially launched CIPHER-SKI (Chemical Intelligent Physics Hybrid Exploration Rewriter – Scaffold Kinetic Insight), a groundbreaking hybrid AI platform designed to accelerate and transform small-molecule drug discovery.
CIPHER-SKI introduces a fundamentally new approach to molecular design. By combining explicit graph rewriting (a technique rooted in lambda calculus and functional programming) with energy-guided Deep Q-Learning and principles from statistical mechanics, the platform achieves true scaffold hopping; the ability to generate structurally novel, patentable chemical matter rather than incremental analogs of existing compounds.
Early validation on the notoriously difficult Switch-II pocket of KRAS G12C (PDB 6OIM), a major oncogenic driver in lung, colorectal, and pancreatic cancers has already yielded impressive results. CIPHER-SKI generated multiple novel scaffolds with strong predicted binding affinities, including mol_000598 at -11.1 kcal/mol, demonstrating excellent shape complementarity in a pocket long considered highly challenging for non-covalent small molecules.
“CIPHER-SKI represents a genuine leap beyond conventional generative AI,” said Gershon Wolfe, Ph.D., Founder of CIPHERICA. “We’re not just optimizing within known chemical space, we’re intelligently rewriting chemistry itself, guided by physics and statistical mechanics, to create better medicines, faster.”
About CIPHER-SKI CIPHER-SKI integrates Graph Neural Networks, energy-driven reinforcement learning, real physics-based feedback (docking + quantum energies + molecular dynamics), and controlled graph transformations to explore chemical space with unprecedented creativity and precision.
CIPHERICA is actively seeking strategic partners and investors to expand platform validation and accelerate the development of novel therapeutics.