Detrended Fluctuation Analysis and Discrete Wavelet Transform for Similarity Comparison of RNA secondary structure

Abstract

Background Ribonucleic acid (RNA) is an important biological macromolecule. Through in-depth studies of RNA, its function has been increasingly discovered. The function of RNA is mostly dependent on its secondary structure because of its conserved nature. The discovery of an approximate relationship between two RNA secondary structures can help to understand their functional relationship better. This discovery can also help in exploring many unknown functions. Currently, RNA secondary structural similarity analysis methods are mainly divided into alignment-based methods and alignment-free methods. Alignment-free methods can obtain similarities and differences among RNA secondary structures more quickly and more accurately than alignment-based methods. Results In this paper, a novel alignment-free method based on the triple vector curve representation of RNA is proposed. A combinational method involving a discrete wavelet transform and detrended fluctuation analysis (DFA) with a sliding window is used to generate the distance matrix. Finally, a phylogenetic tree is constructed using the distance matrix. Experiments are performed on RNA viruses and non-coding RNA datasets, and the phylogenetic trees generated by different methods are compared. The results show that our method yields more accurate results in the comparison of RNA secondary structures. Conclusion The method in this paper enables a more accurate analysis of the similarities between RNA secondary structures. This method has certain application value in the comparison of RNA secondary structures, especially in the analysis of longer RNA sequences.