Prediction of RNA Secondary Structures with XNAfold

Yanming Zou1, Alan J. Hillier2, P. Scott Chandry
1y.zou@pgrad.unimelb.edu.au, University of Melbourne; 2Alan.Hillier@foodscience.afisc.csiro.au, University of Melbourne

Knowledge of the structural features of RNA molecules helps to explain their biological functionality. XNAfold is a GUI application that can be used to predict RNA secondary structure. It is a JAVA-C hybrid program that uses a Java Application Programming Interface and the C-based RNA folding routines in the ViennaRNA library. The program takes an RNA sequence as input and predicts secondary structure using free energy minimization based on experimentally determined thermal dynamic parameters.

The accuracy of the XNAfold predictions was assessed by comparing the predicted structures with 133 experimentally determined structures of RNA available in the Protein Data Bank. The minimum free energy structure predicted by XNAfold was identical to the experimentally determined structure in 58% of cases. Additionally, the experimentally determined structures matched one of the sub-optimal structures computed by XNAfold in 26% of cases. In 9% of cases, the experimentally determined structures did not match any of XNAfold predicted structures. XNAfold cannot predict pseudoknot structures, which were present 7% of the experimentally determined structures.

XNAfold provides an easy to use interface that displays the predicted RNA secondary structure for a given RNA sequence. It is freely available at http://www.student.unimelb.edu.au/~yanmz/index.html