[Thesis Defense] Biomolecular Modeling at Scale

Predicting the structure and interactions of biomolecules is a key challenge in computational biology with broad implications for drug discovery and therapeutic design. Recent advances in deep learning have enabled significant progress in modeling complex molecular systems, but challenges remain in scaling these methods and applying them across diverse biological contexts. The work in this thesis introduces deep learning approaches for biomolecular modeling at scale, with a focus on efficiency, generalizability, and accessibility. I begin by presenting models designed for the general molecular domain, including methods for structure prediction and molecular interaction modeling across proteins, nucleic acids, and small molecules. Building on this foundation, I present applications of these tools in immunology, specifically to the problem of T-cell receptor recognition. Together, these contributions provide a framework for biomolecular modeling that bridges foundational methods and domain-specific applications.