Events Calendar

Improving RNA secondary structure prediction

Speaker: Michelle (Shel) Swensen, Postdoctoral Research Associate

Talk Title: Improving RNA secondary structure prediction

Abstract: RNA folding is one of the fundamental open problems in computational molecular biology. Thermodynamic optimization approaches, which find structures with minimum free energy (MFE), remain the most widely used RNA secondary structure prediction methods. Though these predictions do not always match known structures, the expectation is that structures with a lower free energy are more likely to contain native base pairings, even when the predicted MFE structure itself is not correct.

In this talk I will discuss two avenues for improving structural prediction in a thermodynamic framework: considering large sets of probable structures and augmenting thermodynamic models with additional experimental data.

The Boltzmann distribution specifies that the probability of RNA secondary structure is proportional to an exponential of the negative of its free energy. We present a novel combinatorial method for identifying patterns in structural elements across a Boltzmann sample. Our approach is based on classifying structures according to features chosen from well-defined structural units called helix classes. We show that this combinatorial profiling is straightforward, stable and surprisingly comprehensive.

Data from recently emerging high-throughput structure probing technologies, such as the SHAPE method, have been used in the framework of thermodynamic optimization to predict RNA secondary structure. Via stochastic simulations, we investigate the factors influencing the accuracy of SHAPE data-directed predictions as well as the potential of auxiliary data to further improve prediction accuracy.

Biography: Shel Swenson's training and research interest position her at the intersection of mathematics, computer science, and biology, where she utilizes discrete mathematics to answer questions in molecular and evolutionary biology. Her dissertation, completed under the advisement of Tandy Warnow at The University of Texas at Austin, developed methods for estimating large-scale evolutionary histories. Dr. Swenson is currently a postdoctoral fellow in the School of Mathematics at the Georgia Institute of Technology where she collaborates with Christine Heitsch and her students on problems in mathematical and computational biology.

Host: Professor Viktor K. Prasanna

Location: Hughes Aircraft Electrical Engineering Center (EEB) - 248

Audiences: Everyone Is Invited

Contact: Janice Thompson