基于盒维数原理计算蛋白质的分形维数

Fractal Analysis of Proteins Based on Box Dimension

蛋白质表面的粗糙性和内部结构的不规则性具有明显的分形特征.依据直观的盒维数原理对PDB库中收录的典型酶蛋白、球蛋白和膜蛋白共72种蛋白质的分形维数进行了简洁地模拟计算.数据表明:蛋白质的分形维数(Df)介于1～3之间;分形方法可定量描述蛋白质分子结构的复杂性,即Df值越大,蛋白结构越复杂;Df与残基数的不同步性证明蛋白质分子并不是简单的肽段堆积,而是存在特定的空间结构;同功能酶蛋白具有相近的Df值表明分维是区别于各种宏观参量而对蛋白质微观结构与功能进行表征的有用工具.

Proteins with rough surfaces and irregular structures are characteristic of fractal nature. Fractal dimensions of a set of 72 proteins including typical enzyme, globulin and membrane protein selected from PDB database were calculated based on the principle of intuitive box dimension. The results showed that fractal dimensions of proteins were between 1 and 3, and the more complicated structure a protein has, the larger value of fractal dimension the protein is, so the fractal theory could be used to quantitatively describe the complexity of proteins structure. The discordance between fractal dimensions and the number of residues demonstrated that proteins not only were composed of peptide chains, but also had the specifically spatial structure. The close Dfvalues of the enzymes with similar function indicated that Df, differed from a variety of macro-parameters, was a useful tool for the characterization of micro-structure and function of proteins.