DescriptionRibonucleic Acid (RNA) is an important cellular macromolecule vital to most if not all life on Earth. RNA has many different roles in the cell, most notably as the intermediary molecule that transfers genetic information from DNA to protein in translation. Recently, additional functions of RNA have been elucidated more clearly, such as catalyzing chemical reactions and regulating gene expression. These exciting new findings have shined a scientific spotlight on the field of RNA structure in order to better understand how the once mundane polynucleotide acts in such myriad ways.
An important factor in RNA’s versatile nature is the inherent variation in its chemical structure. The hydroxyl group present on the ribose sugar of a ribonucleic acid makes the corresponding polynucleotide capable of chemical reaction, with itself or with other molecules in the cell. This hyper-reactivity allows RNA to form substantially unique structures, from the hammerhead ribozyme's helical shape from which it takes its name, to the L-shaped conformation common to all transfer RNAs. The problem at hand is thus to study RNA structure and determine if any new patterns can be discovered.
The work presented here centered on a collaborative effort to define a set of conformations common to two-nucleotide long sequences of RNA found in structures from the Protein Data Bank (PDB). This work contributed by clustering RNA di-nucleotides by their torsion angle space using a Fast Fourier averaging technique proven to be effective in clustering nucleotide structure. Each group in the collaboration used different methodologies to analyze the same RNA structural data, and yet found similar results. The collaboration ultimately produced a set of 46 consensus conformations defined by the seven dihedral angles of the sugar-to-sugar unit in a di-nucleotide RNA sequence.
To utilize this new set of RNA di-nucleotide conformations, a software tool was designed and developed to automatically assign the conformation nomenclature to input RNA structure. The program was successfully tested on the pilot study data. A test study was performed on a unique set of RNA structures. The results of this study demonstrated that the consensus conformation set can in fact be used to classify RNA structure.