The Research On Repressntation Of RNA Secondary Structure And Similarity Analysis

RNA is one of the most significant biological macromolecules, it plays extremely important role in the whole life process. So to discover and to define the function of RNA have been an important subject in view of the structure, the similarity comparison of RNA secondary structures has been one of the hotpots in these fields.As one of the major subject in the field of bioinformatics, the similarity comparison of RNA secondary structures own more important research value and a wide field of application with good prospects in the field of function-discovering of RNA. However, because of the differences between representations of RNA secondary structures and abstractions for the representations, methods of RNA secondary structure comparison are varied. Recently, the most commonly representations used in analysis of similarity between RNA secondary structures are classified into two groups, geometry representation and tree model. The first one is known as multi-dimensional representation, which translates RNA bases to points in multi-dimensional space through functions, using constraint between points analysis the similarity between RNA secondary structures. The result of this method is just a number lacking elicitation information for RNA secondary structures. And this representation has some deterioration. The other representation is tree-model-based and already has a wide application in RNA secondary structure comparison.A classic algorithm for comparing trees is presented in details in this paper, and in the later section of the paper, we made a detailed analysis of one kind of tree models and designed an algorithm to compare two trees as the basis of comparing two RNA secondary structures. The main work and innovation of this paper are as follows:1. Made a detailed analysis of tree models of RNA secondary structure representation, and put forward an abstraction of one kind of tree models, vector representation, to describe the planar structure of tree.2. Presented an algorithm based on dynamic programming to find the max sub-structure of two trees. Unlike common algorithms of comparing trees by editing trees, we complete our algorithms by analyzing the consistency of two vectors.3. As the common predicting result of RNA secondary structure is presented in brackets plot, but not in the format of comparing trees. So this paper proposed a validate algorithm to covert from brackets plot to the vector of comparing tree. 4. Analyzed similarity of the structures of influenza virus which are isolated from different areas and hosts.