Core-shell potential derived charges: A detailed analysis

COMP 221

Jeffrey S. Tan, jeffstan@pharmacy.purdue.edu1, Raymond P. Scaringe2, and Kenneth R. Morris1. (1) Department of Industrial and Physical Pharmacy, Purdue University, 575 Stadium Mall Drive, West Lafayette, IN 47907, (2) Pharmaceutical Research Institute, Bristol-Myers Squibb
A core-shell potential-derived (CSPD) charge model is presented for the purposes of modeling electronic environments in molecular assemblies. Each atom is decomposed into a core with fixed charge and position representing the atomic nucleus, and a dynamic shell with its charge and position as the model variables. These variables are optimized only to emphasize reproduction of a quantum mechanical electrostatic potential, as defined in a least-squares-based error function. Point charges are obtained from a modified singular value decomposition algorithm while charge positions are obtained from a conjugate gradient algorithm. This non-linear system requires iterative solutions for the two sets of variables until convergence. Comparisons between traditional potential derived point charge models and CSPD will be discussed. The effects of electrostatic potential sampling on the characteristics of the charge model will also be discussed.
 

Poster Session
6:00 PM-8:00 PM, Tuesday, 12 September 2006 Moscone Center -- Hall D, Poster

Sci-Mix
8:00 PM-10:00 PM, Monday, 11 September 2006 Moscone Center -- Hall D, Sci-Mix

Division of Computers in Chemistry

The 232nd ACS National Meeting, San Francisco, CA, September 10-14, 2006