Accomodating Statistical and Physics-based Computational Protocals for Molecular Glue Design and Evaluation

Molecular glues (MGs) offer a promising strategy for modulating protein-protein interactions (PPIs) by inducing novel intermolecular contacts or stabilizing existing ones. However, the rational design of MGs is hampered by the challenge of predicting the structural arrangement of the ternary complex (MG-protein1-protein2). Here, we present a computational methodology for constructing ternary complexes of MGs, circumventing the need for experimentally determined structures. Recognizing the diverse mechanisms of MG action, we employ two distinct computational approaches tailored to specific MG systems. We validate our methodology through case studies of known MGs, demonstrating its ability to capture the fundamental principles governing MG activity. Our findings highlight the potential of integrating statistical and physics-based methods for the discovery and development of novel MGs.