This workshop focuses on the complexity of dynamics and kinetics in complex systems such as proteins in terms of the underlying multidimensional energy landscape and network structure in conformational and, in general, state space. We aim at having an interdisciplinary meeting that brings researchers together from different fields such as time series analysis, energy landscape, network analysis, and single molecule biophysics.

This year, in addition to the new exploration along the perspectives of energy landscapes, and networks in biological systems, we address the related subjects of
(1) How can we extract the underlying multidimensional manifolds or structures from limited information such as one or a few observable(s)? If people could access the detailed information of molecular dynamics in terms of computer simulation in the era of Boltzmann and Gibbs, people might have devoted most of their time to reach the ergodic hypothesis in the stream of the actual multivariate data, and even might not have come up with the idea of constructing statistical mechanics. In a sense the development of single molecule measurements might have
"shut down" some possible pathways in science. However, it has certainly provided us with a new level of
understanding of what a system actually "feels" along the process of an event, something which has been masked so far in ensemble measurements. Nonlinear time series analysis to elucidate the multidimensional manifolds or structures from scalar time series should become crucial for this problem: One of the questions to be addressed is
how we can extract the relevant information of the
underlying energy landscape and state-to-state network in many-dimension solely from one or a few observable(s) such as time series of the distance between two dye molecules in proteins.
(2) How can we extract the meaningful low dimensional structure buried in the many dimensions from the stream of the multivariate data? In the stream of the actual
multivariate time-dependent/independent data such as molecular dynamics simulation and genome informatics, to extract the meaningful low dimensional structure from them is one of the most crucial problems in many fields. Principal component analysis is one of the traditional approaches. For instance it has been known that the projection of multidimensional free energy landscapes onto a single coordinate such as the number of native contact misleads us regarding the relevant topology the system actually "sees" in the conformation space. One of the questions to be addressed is how much one can reduce
the dimension of the space without losing the important information of the complexity of the system. For multidimensional energy landscapes, the state-to-state network and (transition) disconnectivity graph approaches can be among the possible candidates while the reduction or simplification of the network should still be desirable.

In this workshop, we aim at stirring/combining several disciplines in the areas of energy landscape, network, data-mining and linear/nonlinear dimensionality reduction for the understanding of the complexity of dynamics and kinetics in biological and other complex systems from
different aspects.