Yale University
Department of Statistics
Seminar

Monday, November 3, 2003

Jin Feng
Department of Mathematics and Statistics
UMass-Amherst

Title:  Relative entropy, large deviation and binding tightness between two molecules

Abstract:  Given a ligand and a protein, we are interested in defining a "computable" score that measures the binding tightness between
them. Such problem plays an important role in molecular level drug designs.

We approach the problem from a physical modeling perspective. We view proteins and ligands as groups of amino acids which are modeled as
multi-particle systems interacting through potential functions defined through a small number and well known types of interactions. In
equilibrium, the system relaxes to Gibbs distributions. One sensible choice for the score is the relative entropy between the ligand-protein
joint system and the ligand-protein as two independent systems. Such score can be interpreted as the amount of "free energy" one need to
supply externally, in order to break a binded ligand-protein system apart into two non-interacting sub-systems. Time permitting, I will also
discuss the non-equilibrium situation by using the large deviation theory for stochastic processes.

This is an on-going joint work with Paul Dupuis.
---------------------------------------------

Seminar to be held in Room 107, 24 Hillhouse at 4:15 pm