Biocomputing Research Group on
„Effective Dynamics of Complex Systems“
Many complex systems exhibit metastable states. They correspond to almost-invariant subsets of the state space of the system or to almost adiabatic subspaces. The effective dynamical behavior of the system can approximately be described by the transition statistics between these metastable states. This defines a discrete Markov process with the dominant metastable sets as macrostates and transition rates as induced by the microsccopic dynamics. This Markovian description of the effective dynamics, known under the names “Markov State Models” (MSMs) or “Kinetic Monte Carlo”, can be constructed for many different types of complex systems including, e.g., biomolecular systems, climate dynamics or systems from materials science.
- construct a complete approximation theory that explains why coarse MSMs often are very good models for the effective dynamics,
- construct efficient algorithms for identification of the dominant metastable states and computation of transition rates and pathways between them,
- use MSMs to understand the effective dynamics on timescales beyond those that are accessible by direct numerical simulation of the dynamics,
- incorporate non-equilibrium perturbations,
- and apply all this to real-world systems.
For project details please click here.
- Sarich, M. and Noé, F. and Schütte, Ch. (2010) On the Approximation Quality of Markov State Models. Multiscale Model. Simul., 8 (4). pp. 1154-1177.
- Noé, F. and Schütte, Ch. and Vanden-Eijnden, E. and Reich, L. and Weikl, T. (2009) Constructing the Full Ensemble of Folding Pathways from Short Off-Equilibrium Simulations. Proc. Natl. Acad. Sci. USA, 106 (45). pp. 19011-19016.