AveRNA

Ensemble-based Prediction of RNA Secondary Structures

(Nima Aghaeepour and Holger Hoos)

Abstract:
Accurate structure prediction methods play an important role for the understanding of RNA function. Energy-based, pseudoknot-free secondary structure prediction is one of the most widely used and versatile approaches, and improved methods for this task have received much attention over the past 5 years. Despite the impressive progress that as been achieved in this area, existing empirical evaluations do not permit a comprehensive, statistically sound assessment of the prediction accuracy achieved by various methods. Furthermore, while there is increasing evidence that there is no single prediction algorithm that consistently outperforms all others, no work has been done to exploit the complementary strengths of multiple approaches. In this work, we address both of these issues. Firstly, we introduce a set of statistical tools that, together with a previously published and increasingly widely used dataset of high-quality RNA structures, provides a framework for the objective evaluation of these methods. Secondly, we introduce AveRNA, a generic and powerful method for combining a set of existing secondary structure prediction procedures into an ensemble-based method that achieves higher prediction accuracies than obtained from any of its component procedures. Our experimental results clarify the performance relationship between 10 well-known existing energy-based pseudoknot-free RNA secondary structure prediction methods and clearly demonstrate the progress that has been achieved over the last 5 years. We furthermore demonstrate that our new, generic ensemble-based procedure, AveRNA achieves significantly higher prediction accuracy than any of the previous algorithms. In addition, AveRNA allows an intuitively and effectively control the trade-off between false negative and false positive base pair predictions. Finally, AveRNA can make use of arbitrary sets of secondary structure prediction procedures and can therefore be used to leverage improvements in prediction accuracy offered by procedures developed in the future.

Resources:

• Web Server - AveRNAs webserver for predicting the structure of an RNA strand (initial version; advanced version under preparation).
• Source code - AveRNAs scripts for: 1) predicting the structure of an RNA strand; 2) reproducing the tables and figures in the manuscript bellow.
• SSTRAND2 Dataset - The full set of RNA structures
• Trainingset - A uniformly sampled subset of SSTRAND2, used for parameter optimization.

Papers:

• Nima Aghaeepour and Holger Hoos:
  Ensemble-based Prediction of RNA Secondary Structures.
  RNA, submitted for publication, May 2011.