Modeling RNA Folding Landscapes with Probabilistic Roadmap Methods

* Montana State University
+ Texas A & M University

RNA molecules, which are composed of a sequence of nucleotides, fold into energetically optimal conformations. These conformations exist in three-dimensions. The properties of these conformations allow the process to be examined in two steps, the first involves folding of secondary structures in two dimensions, while the last step expands the folding to the tertiary structure and the third dimension. We describe the kinetics of the folding process using a folding landscape that depicts possible conformations and their energy values. Folding from one conformation to another proceeds along the lines described by the landscape, similar to the way a ball rolls down a hill.

The particular problem we are interested in is that of describing the most significant features of the landscape. The landscape is too large to explore completely, so we are using the Probabilistic Roadmap Method (PRM) to find the conformations that have the greatest impact of the folding process. Once we have described the landscape, we use a differential equation to describe the population kinetics of the RNA. Our methods allows us to discover features such as folding rates, transition states, and folding pathways. Previously, we successfully applied this technique to protein folding landscapes.

Presentations
USRG Symposium Presentation TAMU, 6 Aug 2003

Paper
B. Kirkpatrick, X. Tang, S. Thomas, N. Amato, "Modeling RNA Folding Landscapes with Probabilistic Roadmap Methods", Technical Report, TR03-004, Texas A&M University, Aug 2003. Technical Report ( ps, pdf )