A major challenge when simulating any peptide system is efficiently sampling the entire structural ensemble of the peptide. Sampling a full structural ensemble proves to be especially difficult for cyclic peptides. Due to their constrained nature, cyclic peptides behave radically different from their linear counterparts, and their free energy profiles often contain large barriers between structural conformations, making structure sampling with traditional molecular dynamics difficult on a reasonable timescale. Thus enhanced sampling techniques, such as metadynamics, have been used to help accelerate conformational sampling. However, with a technique such as metadynamics an <i>a priori</i> knowledge of which slow degrees of freedom the system is struggling to overcome is crucial to efficiently enhance sampling. In our research, we aim to gain knowledge of how model cyclic peptides change conformations so that we may develop new enhanced sampling techniques to help alleviate the conformational sampling problem. We first simulated a relatively flexible cyclic hexapeptide using both conventional molecular dynamics and metadynamics. By studying the kinetics and thermodynamics of the system we found that conformational switches occur via coupled two-dihedral angle changes. Using this knowledge we developed a new set of collective variables to target these coupled two-dihedral angle changes, which greatly reduces the simulation time necessary to sample the entire structural ensemble of a cyclic peptide.

Developing enhanced sampling methods for cyclic peptides

253rd National Meeting (2017)

The official archive and repository for all ACS Meetings & Expositions abstracts and select posters or presentations.