β-shape을 이용한 단백질의 구조 정렬

Abstract

A protein is a large organic compound which plays an important role in a living cell. It is known that a structure of a protein is one of the important factors which determines its function. Therefore, it is important to find proteins with similar structure. For this purpose, the proteins are required to be appropriately aligned so that their structures are maximally superimposed. Structure alignment problem is to locate the correspondence pairs between proteins (the equivalence set) which yields the maximal superimposition. Once the protein structures are aligned, the biological problem such as protein structure similarity and docking can be facilitated. One of the previous approaches to structure alignment problem is to alternately superimpose the two structures and to find the equivalence set. In this paper, we propose a new approach to protein structure alignment based on the β-shapes of proteins. The previous approach used the only sequence information to determine the equivalence set. However, the proposed method exploits the shape information of a protein to locate the equivalence set via a β-shape. Since a β-shape of a protein contains unique properties of the protein, the structure alignment can be more effectively via β-shapes. Experiments on the comparison with previous approach are performed for several sample proteins from Protein Data Bank (PDB).