(572e) Linking Atomistic and Mesoscales: Atomistic and Simple Coarse-Grained Models in Application to Biomolecular Problems

Computer simulations and molecular modeling provide powerful tools to underpin and explain phenomena occurring on microscopic time and length scales (nanometers, nanoseconds). However for many systems, and in particularly for biological systems, we are interested in effects involving larger scales and taking place over longer periods of times (microseconds). In these situations, simplified, coarse-grained models provide a useful route. The idea of coarse-grained models is to reduce the number of degrees of freedom in the system and thus to make it more computationally tractable. Although, in many cases these models are qualitative, they may provide fundamental insights on the phenomena otherwise unattainable.

In this presentation we will consider several examples of atomistic and coarse-grained models in application to biomolecular problems. The first example considers a particular class of peptides capable of seamless translocation through cell membranes. Due to this property these peptides have been extensively investigated as potential vectors for cellular drug delivery. The actual mechanisms of translocation however vary from system to system and in many cases are unknown. We will construct simple models of these peptides to elucidate these mechanisms and discuss the implications of the model.

In the second example, we employ non-equilibrium molecular dynamics simulation to characterize the effective interactions between lysozyme molecules involved in the formation of two hydrophobic crystal contacts. We show that the effective interactions between crystal contacts do not exceed a few kTs, the range of the attractive part of the potential is less than 4Å and that, within this range, there is a significant depletion of water density between two protein contacts. Our findings highlight the different nature of protein crystallization and protein recognition processes.