University of Kentucky
Date & Time
February 16, 2024, 2:00 pm – 3:00 pm
Title: Emerging Mathematics and Geometric Deep Learning for Drug Design
Abstract: The intersection of mathematics and artificial intelligence (AI) has ushered in a new era of drug design, offering unprecedented accuracy and efficiency in identifying potential drug candidates. This talk focuses on mathematics representation learning and geometric deep learning, rapidly advancing areas within machine learning and data mining that specialize in processing graph-structured and 3D data. Our exploration includes the latest developments in differential geometry, persistent spectral graphs enhanced with geometric deep learning, and advanced large language models. These tools have proven instrumental in characterizing biomolecular and molecular interactions. A standout feature of our approach is its scalability, which accommodates diverse molecular representations, and its robustness, especially when handling low-quality data. These strengths have catapulted our models to the forefront, as evidenced by our top-tier performance in the small molecular properties benchmarks, protein-protein interactions, and the D3R grand challenges. This presentation will offer insights into the future trajectory of drug design, where the synergy of mathematics and AI is set to redefine the boundaries of what is achievable in pharmaceutical research.