Despite the biological importance of non-coding RNA, their structural characterization remains challenging. Making use of the rapidly growing sequence databases, we analyze nucleotide coevolution across homologous sequences via Direct-Coupling Analysis (DCA) to detect nucleotide-nucleotide contacts. DCA is a global statistical modeling approach, which has been originally developed for predicting residue-residue contacts in proteins, and which has been used successfully for guiding tertiary and quaternary protein structure prediction.

In the case of RNA, the use of coevolutionary information has been restricted to secondary structure prediction. For a representative set of riboswitches, we show that the results of DCA in combination with a generalized Nussinov algorithm systematically improve the results of RNA secondary structure prediction beyond the traditional covariance approaches based on mutual information. Even more importantly, we show that the results of DCA are enriched in tertiary structure contacts. By integrating these predictions into molecular modeling tools, systematically improved tertiary structure predictions can be obtained, as compared to using secondary structure information alone.