Conformation dynamics is a recent concept suggested and established by the author and Christof Schuette. The key to this concept is the direct identification of metastable conformations together with their life times and transition patterns. Once a certain Markov operator has been discretized, a stochastic matrix arises, which can be treated by Perron cluster (cluster) analysis, meanwhile called PCCA. The method involves the numerical solution of a stochastic eigenproblem for the Perron cluster of eigenvalues and analyzes the problem in terms of sign patterns of the components of the Perron cluster eigenvectors. As has turned out in extensive applications to drug design problems, some undesirable discontinuity and lack of robustness enters into the PCCA algorithm. The talk will present a more robust variant of Perron cluster analysis called
PCCA+, which has been worked out in detail very recently by the
author and Marcus Weber. The new approach analyzes the eigenvector data in a rather different setting, partly using techniques of discrete optimization. Numerical examples of biomolecules including HIV and SARS will be included.
Copyright. All Rights Reserved.