Mechanistic Dissection of Conformational Transition of Bicyclic Peptide via Molecular Modeling and Deep Learning.

Molecular conformations play a critical role in determining molecular properties, such as membrane permeability, binding affinity, and ultimately therapeutic efficacy. Experimental and computational approaches can characterize conformations and provide insight into why certain conformations are thermodynamically preferred over others. However, examining conformations alone may not fully explain why subtle differences, such as a single LEU-to-ILE mutation in a bicyclic peptide, can produce markedly distinct conformational ensembles. Analyzing the transition pathways between conformations further reveals the mechanisms that shape these ensembles. Here, we introduce a deep learning model, termed ICoN-v1, trained in molecular dynamics simulation data to learn the underlying physics that governs cyclic peptide conformational dynamics. We examined hexacyclic peptides with Nuclear magnetic resonance (NMR)-determined structures, and MYC-targeting bicyclic peptides, which are stereo-diversified or have a single LEU-to-ILE mutation. By following the minimum-energy pathway in the latent space constructed by ICoN-v1, conformational transition paths, led by various sets of concerted backbone and sidechain torsional rotations moving in sequence between energy minima, are efficiently generated. Notably, smooth transition pathways that are absent from molecular dynamic output can be observed using ICoN-v1. Our results identify various sets of concerted torsional motions that are nonlinearly combined during conformational transitions and reveal the key residues governing each stage of the transition, thereby elucidating how the observed conformations are generated and informing molecular design.