TARNAS: A Software Tool for Abstracting and Translating RNA Secondary Structures

Abstract

Ribonucleic acids (RNAs) fold into complex structures that are strongly associated with their biological functions. These can be abstracted into secondary structures, represented as nucleotide sequences annotated with base-pairing information. This abstraction is both biologically relevant and computationally manageable. Comparing and classifying RNA molecules typically relies on these secondary structure representations, which exist in multiple formats. In this work, we introduce TARNAS 1.0, a software tool designed to convert RNA secondary structure representations across multiple formats, including Base Pair Sequence (BPSEQ), Connect Table (CT), dot-bracket, Arc-Annotated Sequence (AAS), Fast-All (FASTA), and RNA Markup Language (RNAML). The tool offers options for retaining or removing comments, blank lines, and headers during the conversion process. These format translation and preprocessing capabilities are specifically designed to support the batch handling of large collections of RNA molecules, making TARNAS well suited for large dataset construction and database curation. Beyond format translation, TARNAS computes three levels of abstraction for RNA secondary structures, namely core, core plus, and shape, as well as a set of statistical descriptors for both primary and secondary structure. These abstraction and analysis features are intended to facilitate the comparison of molecules and the identification of recurring structural patterns, which are essential steps for associating structural motifs with molecular function. TARNAS is available as both a standalone desktop application and a web-based tool. The desktop version supports batch processing of large datasets, while the web version is optimized for the analysis of single molecules.

Keywords: RNA secondary structure formats; RNAML; abstractions of RNA secondary structures; statistics on RNA secondary structures.

Figure 1

Arc diagram representation of RNA secondary structures. Red circles indicate nucleotides, while arcs denote base-pair interactions. The motif shown in part (a) is pseudoknot-free, whereas the one in part (b) contains a pseudoknot.

Figure 2

The shape (a), core (b), and core plus (c) abstractions of the RNA secondary structure depicted in Figure 1.

Figure 3

Interface of the TARNAS web app.

Figure 4

Interface of the TARNAS standalone version.

Figure 5

BPSEQ (a) and CT (b) representations of the RNA secondary structure from Figure 1.

Figure 6

RNAML (a), AAS (b), dot-bracket (c), and FASTA (d) representations of the RNA secondary structure from Figure 1.

Figure 7

Schematic view of key RNAML markup elements. (a) Top-level RNAML elements. (b) Elements within the molecule description. (c) Elements used in tertiary structure annotation.

Figure 8

TARNAS workflow for translation. Each molecule file is parsed using either ANTLR4 or the W3C DOM, producing an internal representation of the secondary structure. This internal structure is then rendered in the requested output format.