Conformational Ensembles from Experimental Data and Computer Simulations

112 / 161

Conformational Ensembles from Experimental Data
and Computer Simulations Poster Abstracts
108
73-POS Board 33
Dimensional Reduction of Markov State Models from Renormalization Group Theory
Simone Orioli1,2, Pietro Faccioli1,2 .
1University of Trento, Trento, Trento, Italy, 2Trento Institute for Fundamental Physics, Trento,
Trento, Italy.
In this work we used Renormalization Group (RG) theory to define a rigorous algorithm to
cluster the microstates of a Markov State Model into macrostates. The result is a lower-
dimensional Master Equation which yields the optimal reduced Markovian description of the
system’s relaxation kinetics. To illustrate and validate our method we analyze a number of test
systems of increasing complexity, ranging from synthetic toy models to atomistic molecular
dynamics smulations. In all cases, the low-dimensional Markov State Model is found to
reproduce the kinetics of the original model with very high accuracy, with a relative error of at
most a few percent.