Computationally efficient matching of MicroRNA shapes using mutual symmetry

Arjan Kuijper,I. Havukkala
2007-08-08
Abstract:In this work we investigate the use of mutual symmetric points in matching microRNA shapes. These 2D shapes, obtained by an RNA folding algorithm, are given as coordinate points of 80--140 nucleotides long microRNA sequences containing A-U and C-G nucleotide pairs. The method of mutual symmetric points, exploiting global aspects, is well-suited for analysing these shapes that contain global symmetries. We discuss how a parameterisation can be derived from the coordinate point representation of these shapes. As this results in a data set with too many points, we investigate how reasonable subsets of this parameterisation can be used to compare microRNA shapes with respect to optimising both the computational effort and the search accuracy and sensitivity. We conclude that taking a uniform step size parameterisation with about 3 times as many points as nucleotides performs best in view of computational efficiency and effectiveness. It should be taken into account whether shapes are aligned or not.
What problem does this paper attempt to address?