Poster abstracts
Poster number 133 submitted by Jeffrey Zuber
A sensitivity analysis of RNA folding nearest neighbor parameters quantifies the uncertainty in RNA secondary structure prediction and reveals correlations between parameter values
Jeffrey Zuber (Biochemistry and Biophysics, University of Rochester), B. Joseph Cabral (Computational Sciences, Moderna Therapeutics), Iain McFadyen (Computational Sciences, Moderna Therapeutics), David M. Mauger (Computational Sciences, Moderna Therapeutics), David H. Mathews (Biochemistry and Biophysics, University of Rochester)
Abstract:
RNA nearest neighbor thermodynamic parameters are often used to predict RNA secondary structures in order to develop hypotheses about structure-function relationships for RNA sequences, in RNA sequence design, and to identify new functional RNAs. These parameters were derived from optical melting experiments on sets of small model oligonucleotides. To better understand the precision of RNA secondary structure prediction, this work examines how experimental errors in optical melting experiments propagates to errors in the derived nearest neighbor parameter values and hence to uncertainty in structure prediction. This work identifies previously undocumented correlations between nearest neighbor parameters. This work also shows that the precision of RNA secondary structure prediction is more robust than was indicated by previous work that was based on perturbation of the nearest neighbor parameters. Additionally, this work provides an estimate for the precision of RNA secondary structure prediction as well as guidance for future optical melting experiments.
Keywords: RNA Secondary Structure, Nearest Neighbor Parameters, Sensitivity Analysis