Capturing alternative secondary structures of RNA by decomposition of base-pairing probabilities

Abstract

BackgroundIt is known that functional RNAs often switch their functions by forming different secondary structures. Popular tools for RNA secondary structures prediction, however, predict the single ‘best’ structures, and do not produce alternative structures. There are bioinformatics tools to predict suboptimal structures, but it is difficult to detect which alternative secondary structures are essential.ResultsWe proposed a new computational method to detect essential alternative secondary structures from RNA sequences by decomposing the base-pairing probability matrix. The decomposition is calculated by a newly implemented software tool, RintW, which efficiently computes the base-pairing probability distributions over the Hamming distance from arbitrary reference secondary structures. The proposed approach has been demonstrated on ROSE element RNA thermometer sequence and Lysine RNA ribo-switch, showing that the proposed approach captures conformational changes in secondary structures.ConclusionsWe have shown that alternative secondary structures are captured by decomposing base-paring probabilities over Hamming distance. Source code is available from http://www.ncRNA.org/RintW.

RintC: fast and accuracy-aware decomposition of distributions of RNA secondary structures with extended logsumexp

“…RintW + maximum-span. At first, we introduced maximum-span constraint in base pairs to the baseline algorithm of RintW (6). Detailed descriptions of RintW and the proposed method are in the accompanying supplementary file.…”

“…The computational costs of basic algorithms used by previous methods are high, but the Fourier transform has been shown to reduce the time complexity of analysis. (4), (5), (6) Those previous methods, however, cannot be used to analyze long ncRNAs (e.g., > 1000 nucleotides) because the computational complexities are still too high. In this paper, we show that further reduction of computational complexity is possible by introducing maximum-span constraint on basepairs (7).…”

RintC: fast and accuracy-aware decomposition of distributions of RNA secondary structures with extended logsumexp

“…We measured the computation time in the case where the maximum-span constraint W = 100 is introduced and in the case where no restriction is applied (equivalent to RintW [7]) ( Fig. 1).…”

“…Initially, we introduced a maximum-span constraint in base-pairs to the baseline algorithm of RintW [7]. Detailed descriptions of RintW and the proposed method are described below.…”

“…Structures of long ncRNAs (e.g., > 1000 bases, including ribosomal RNAs) are important for understanding their functions, but analyzing the probability landscape of the structure remains a challenging task. Fourier transform has been utilized to reduce computational complexity [5][6][7], but the computational costs of previous methods are still too high to apply to long ncRNAs. Furthermore, the Fourier transform distributes numerical errors uniformly across large and small marginal probabilities, which makes small marginal probabilities unreliable.…”