Pairwise local structural alignment of RNA sequences with sequence similarity less than 40%

Area of Science:

• Bioinformatics
• Computational Biology
• Genomics

Background:

• Identifying non-coding RNA (ncRNA) genes and structural RNA elements (eleRNA) is challenging due to conservation in structure rather than sequence.
• Existing methods struggle with pairwise detection of genes exhibiting low sequence similarity within larger genomic regions.

Purpose of the Study:

• To develop and evaluate a novel pairwise local alignment approach for detecting RNA structural motifs.
• To improve the efficiency and accuracy of identifying ncRNAs and eleRNAs with low sequence similarity.

Main Methods:

• The study presents an enhanced foldalign implementation based on the Sankoff algorithm for simultaneous structural alignment.
• The method allows mutual scans of arbitrary length sequences to find common local structural motifs.
• A scoring scheme incorporating structural parameters and substitution matrices (similar to RIBOSUM) was utilized.

Main Results:

• The foldalign method demonstrated comparable performance to Dynalign in identifying common structures but was substantially faster.
• Structure prediction performance reached a Matthews correlation coefficient of approximately 0.7 for RNA families.
• In genomic context analysis, foldalign achieved an average sensitivity of 0.8 and a positive predictive value of 0.9 for locating ncRNAs and eleRNAs.

Conclusions:

• The enhanced foldalign algorithm offers a computationally efficient and accurate solution for identifying structurally conserved RNA elements with low sequence similarity.
• The method is effective in both isolated structural motif detection and localization within complex genomic sequences.
• This tool aids in the discovery of novel ncRNA genes and structural RNA elements, advancing genomic research.