Bridging AI and Molecular Sciences for Next-Generation Drug Discovery

A central challenge in AI-driven molecular science lies in efficiently representing molecular data and developing learning architectures that capture intrinsic structure-function relationships. We introduce advanced mathematics-based molecular representations and learning frameworks. Molecular structures and interactions are encoded using high-order topological and algebraic representations, including Rips complexes, Alpha complexes, Neighborhood complexes, Dowker complexes, Hom-complexes, Tor-algebras, Rhomboid tilings, Sheaves, and Categories. Building on these foundations, we design physics-informed geometric and topological deep learning models that systematically integrate high-order, multiscale, and periodic information of molecular systems.

Molecule discovery is increasingly constrained not by ideas, but by the speed and reproducibility with which chemical matter can be designed, made, and iterated. This seminar will describe a molecular-science-first approach to “bridging AI and Chemistry,” where mechanistic insight, modern synthesis, and automation are integrated into closed-loop workflows that generate both molecules and the high-quality data required to predict the next ones. I will highlight how new reactivity platforms in synthetic methodology, including selective carbonyl olefination and skeletal editing strategies, can convert readily available starting materials into complex, drug-like architectures with rapid diversification at late stages. These transformations provide an ideal testbed for AI, because they expose interpretable structure–reactivity patterns that can be learned, stress-tested, and deployed for design.
I will then discuss how robotic synthesis and high-throughput experimentation enable self-driving optimization, producing reproducible datasets that connect reaction outcomes to controllable variables (substrate features, microenvironment, catalyst networks, and process parameters). Finally, I will outline a translational pathway that couples AI-enabled synthesis to real therapeutic needs through industry partnerships, enabling faster lead generation, improved manufacturability, and more sustainable chemistry. The overarching message is that the next generation of drug discovery will be built not only on better models, but on better molecular platforms that make chemical space accessible, predictable, and scalable.