Replica Exchange with Solute Tempering for Protein Conformational Sampling

Molecular dynamics (MD) simulations have become indispensable in numerous biological studies, facilitated by advances in algorithms and high-performance computing. These simulations have played a pivotal role in uncovering both the structural and functional characteristics of biomolecules, particularly proteins, through sophisticated molecular modeling techniques. However, the comprehensive understanding of molecular functions hinges on the ability to thoroughly explore their conformational space. Despite the power of MD simulations, complex systems such as proteins often present challenges, as they can become trapped in local energy minima within the rugged free energy landscape. While recent advancements in hardware and distributed computing platforms have enabled longer simulations and the study of protein folding in explicit solvents, conventional MD simulations still face difficulties in adequately sampling protein configuration space due to this trapping phenomenon. To address this issue, generalized ensemble (GE) sampling algorithms like the replica exchange method (REM) and simulated tempering (ST) have been developed. These algorithms facilitate the exploration of conformational space by inducing a random walk in temperature space within an expanded ensemble. By enabling the exchange of conformations between different temperature replicas, these algorithms can help facilitate the escape from local energy minima, particularly at higher temperatures where thermal fluctuations are more pronounced. These temperature-based GE algorithms are widely employed in simulating proteins and other biological macromolecules. In this chapter, we provide an overview of replica exchange methods, followed by a detailed discussion of Hamiltonian-based replica exchange methods, with a particular focus on replica exchange with solute tempering (REST), REST2 and their variants, including generalized REST (gREST) and replica exchange with hybrid tempering (REHT). Finally, we explore several applications of these enhanced sampling methods in elucidating biological conformational changes.