Effective All-Atom Potentials for Proteins
Experimental developments are steadily improving our knowledge of the dynamics and interactions of proteins, knowledge that is important for understanding living systems at the molecular level. Computer simulations provide a complementary tool in this endeavor and offer unique opportunities to elucidate salient features that remain experimentally inaccessible. Simulations are being used to study many different aspects of protein dynamics of varying computational complexity, which makes it necessary to choose models for the calculations with care, depending on the problem at hand. The models in use today range from reduced models with one or a few sites per amino acid to all-atom models with explicit solvent molecules. All-atom simulations have traditionally often been based on quite elaborate potentials. This level of detail may be needed in many applications, like structure refinement, but is not an obvious choice when studying processes like folding and aggregation. Some recent (implicit solvent) models use simpler and computationally more convenient potentials, while retaining an all-atom description of the protein chains. This article discusses some potentials of this type. The usefulness of the approach is illustrated by briefly describing studies, based on one of the potentials, of folding thermodynamics, mechanical unfolding and aggregation.
- Frenkel D, Smit B (1996) Understanding molecular simulations: from algorithms to applications. Academic San Diego, CAGoogle Scholar
- Meinke J, Hansmann UHE (2009) Free-energy driven folding and thermodynamics of the 67-residue protein GSαW –a large-scale Monte Carlo study. J Comput Chem 30:1642–1648Google Scholar
- Rapaport DC (1997) The art of molecular dynamics simulations. Cambridge University Press, CambridgeGoogle Scholar