Defensio Abstract

Speaker	Michael Wolfinger
Title	Energy Landscapes of Biopolymers

Biomolecules like DNA, RNA or proteins form the molecular basis of all known forms of life. The ability of biopolymers to fold into a well-defined native state is a prerequisite for biologically functional molecules. In order to treat biomolecules within a theoretical framework, a reasonable level of abstraction or coarse-graining is needed. RNA can be modeled conveniently by means of secondary structures. Based upon experimentally measured energy parameters, efficient dynamic programming algorithms for a computational treatment of RNA secondary structures have been developed at our institute and made available as the Vienna RNA Package. Proteins are often modeled as self-avoiding walks on various lattices with a sequence consisting of only two monomer types, hydrophobic H and polar P residues.

A fundamental prerequisite in complexity studies of molecular systems is certainly a thorough investigation of the energy surface on which the system dynamics evolve. A detailed understanding of structural features of complex landscapes thus lies at the heart of the biophysics of heteropolymers. Kinetics and structure formation processes of biopolymers are crucially determined by the topological details of the energy landscape, i.e. basins and barriers separating them.

We introduce an efficient algorithm for measurement of features of energy landscapes, such as the number of local minima, the size distribution of basins of attraction or thermodynamic quantities. The algorithm is capable of constructing a hierarchical order of conformations that can be represented compactly in so called barrier trees, giving an impression of the shape and ruggedness of the energy landscape.

A stochastic algorithm for the simulation of kinetic folding of RNA, based on elementary steps in conformation has been extended to the field of lattice proteins. We compare results from an extended Arrhenius-type macrostate kinetics, that can be formulated on the barrier tree with results from the stochastic simulation. A major advantage of the coarse-grained dynamics is time efficiency, allowing computational treatment of tRNA size molecules' dynamics within a time-scale of several minutes.

We will further present a novel approach to generate the lowest-energy part of lattice protein energy landscapes based on elementary steps starting from a low-energy state.