A Sampling Approach for the Exploration of Biopolymer Energy Land- scapes

The folding of biopolymers is crucially determined by the properties and the topology of the underlying energy landscape. A reduced representation of these energy landscapes is provided by barrier trees, which can be used to study the dynamical behavior of the folding. We pre-sented a generic, problem-independent approach for the generation of barrier trees of discrete biopolymer models. In contrast to previous studies, the approach used does not rely on enu-meration, which is limited to smaller molecules due to the amount of available memory. The algorithm has been applied to RNA and a lattice protein. The results show that the approach can be used to compute both all local minima and the exact barrier tree of an energy land-scape. The presented method does not restrict the investigated conformation space to certain regions. 1