Kinetic Transition Network Based on Trajectory Mapping

The understanding of biomolecular systems can be greatly improved by constructing conformational transition networks. Network representation honestly reflects the intrinsic complexity of high-dimensional systems, which facilitates the estimation of various thermodynamic and kinetic properties. In this paper, we introduce the trajectory mapping (TM) method for naturally detecting the metastable states of biomolecules and for subsequently constructing the hierarchical kinetic transition networks. In TM, multiple simulation trajectories are mapped to high-dimensional vectors, and the interrelation between these trajectory-mapped vectors is analyzed to locate metastable states. Kinetic information can be quantitatively extracted through simple algebraic manipulations of the identified metastable states. We apply the TM method to a toy model and alanine dodecapeptide. The polypeptide system is analyzed at two time scales, and the metastable states and interstate transition kinetics are correctly revealed.