RNA Secondary and Tertiary Structure Prediction by Tracing Nucleotide Co-Evolution with Direct Coupling Analysis

Abstract

Structural characterization of many important noncoding RNA- typically preceding any detailed mechanistic exploration of their function- remains challenging. In recent years, the field of protein structure prediction has developed a reliable computational alternative. Due to the explosive growth of sequential databases and novel statistical analysis tools it is now possible to trace the co-evolution of amino acids and predict tertiary contacts [1]. These contacts can be exploited as spatial constraints in structure prediction workflows leading to excellent quality predictions with little error [2,3] and effectively revolutionized the field of protein structure prediction.Here, we have successfully adapted one of the premier methods, Direct Coupling Analysis [1], to the specifics of RNA. While Watson-Crick base related to secondary structure are relatively easily predicted, predicting tertiary contacts has remained a challenging task met with limited success. For a representative set of riboswitches we are able to predict many of their tertiary contacts with little error by statistically analyzing their multiple sequence alignments for nucleotide co-evolution [4]. Although riboswitches are considered to be a hard task for RNA structure prediction, we further demonstrate that these tertiary contacts are sufficient to systematically and robustly improve tertiary RNA prediction quality [4]. Considering the large gap of known ncRNA sequences to experimentally resolved tertiary structures, we are convinced that this represents a true breakthrough that significantly supports all RNA related research.