Activity Number:
|
495
|
Type:
|
Invited
|
Date/Time:
|
Wednesday, August 5, 2009 : 2:00 PM to 3:50 PM
|
Sponsor:
|
Section on Statistical Computing
|
Abstract - #302905 |
Title:
|
Auxiliary Variable MCMC with Applications in Protein Structure Modeling
|
Author(s):
|
Jun S. Liu*+ and Kou X. Sam
|
Companies:
|
Harvard University and Harvard University
|
Address:
|
1 Oxford Street, Cambridge, MA, 02138,
|
Keywords:
|
multiple-try ; HP model ; Protein structure ; side-chain entropy ; sequential Monte Carlo
|
Abstract:
|
I will describe some of our recent efforts in the development of Monte Carlo strategies along the idea of multiple-try Metropolis methods. I will illustrate these ideas using examples from Hydrophobic-Hydrophilic (HP) protein model optimization and protein side-chain entropy (SCE) estimation. By applying the new MCMC schemes, we were able to achieve the best results for all the 2-D and 3-D HP structural optimization examples we can find in the literature. We also found that widely used pairwise potential functions behaved surprisingly badly for stabilizing near-native protein structures, and adding a term representing the SCE of the protein can help greatly in discriminating true native structures from decoys. The general MCMC methodology borrows ideas in sequential Monte Carlo and the multiple-try proposal approaches. Joint with J Zhang and S Kou.
|