Detecting and assessing conserved stems for building structural alignment of RNA sequences

Abstract

The comparative methods for identifying non-coding RNAs or predicting RNA secondary structure can be facilitated by taking structural alignments of homologous sequences as input. However, it is very difficult to build a well structural alignment without knowing the secondary structures. Here, we propose a new method for building structural alignment of RNA sequences by detecting and assessing conserved stems across the sequences. The method can be summarised by: we detect conserved stems across multiple RNA sequences using the so-called position matrix with which some common paired positions are uncovered; we assess the conserved stems using the Signal-to-Noise and the profile SCFG; we build the structural alignment by incorporating some compatible conserved stems with the initial sequence alignment built by Clustal W. Experimental results show that our method can build structural alignment of RNA sequences with much greater sensitivity and specificity than Clustal W.