NONLINEAR INDICES ANALYZING METHOD FOR MOLECULAR DYNAMICS OF PROTEIN DYNAMICAL TRANSITION
Abstract
We introduce four kinds of nonlinear indices into the analysis of free energy data produced by protein simulation with molecular dynamics. They are fractional dimension, approximate entropy, Poincare plot indices and predictability index with neural network. In this paper, the 200 K dynamical transition of myoglobin is used as an example, and the free energy data of the simulation is taken as time series to calculate the nonlinear indices. We show that all these indices as a function of temperature allow to indicate the dynamical transition with an obvious turning point, a "valley" or a "peak" around 200 K.
You currently do not have access to the full text article. |
---|