Current Members

Hongtao Yu, SWR, Alexis Lee

American Chemical Society Meeting, New Orleans, 2008

July, 2009: Tucker Adams, David Brooks, SWR, AJL, HY, James Crouch

 

Current Projects

Our research uses theoretical and computational approaches, primarily molecular dynamics and Monte Carlo methods. We have been involved in the development of potential models for water as well as the development of more efficient simulation methods. We use these approaches to study a variety of different systems, including proteins, aqueous solutions, and ice. Two of our active research areas are described below.

Water in confined environments and its influence on ligand binding, assembly, and protein function. Water in protein interiors and at protein-ligand interfaces can influence protein function and ligand binding. For instance, the addition of a single water molecule can apparently change ligand binding constants by an order of magnitude. We are characterizing the thermodynamics in terms of free energy, entropy and heat capacity changes for the addition of water molecules to interior sites with various volumes and varying hydrophobic/ hydrophilic character. Another study involves the nature of water in concave hydrophobic environments as are found in deep cavity cavitand molecules. At right is a 20 picosecond trajectory showing the dynamical nature of the interior water molecules, which vary from 1 to 4 in this trajectory.

 

Protein conformational changes. Many of the systems we study contain large energy barriers. One simulation method for overcoming these barriers is replica exchange (RE), which couples the simulation with higher temperature replicas. The high temperature replicas allow for faster transitions over the activation barriers. One serious drawback of RE is that it scales poorly with system size, so that for even small proteins many replicas are needed. We developed a general, easy to implement RE method that can be used for large systems (replica exchange with dynamical scaling, or REDS). The method requires fewer replicas (for example, 5 rather than 22), greatly increasing its efficiency, as shown in the figure on the right. The errors in the energy of the system decrease much faster with 5 replica REDS than with 22 replica RE. We are currently optimizing the method and applying this method to study the folding of proteins.

4/08