依据氨基酸残基的相关性预测蛋白质的结构类型

作为蛋白质的建筑构件,各种类型的蛋白质的20种氨基酸残基之间存在着特定的相互关联,反映了氨基酸残基之间的制约性,并有深刻的物理和化学的内在因素.某些氨基酸残基对之间的相关系数可以作为一种类型的蛋白质区别于其它类型蛋白质的特征,用于蛋白质结构类型的预测.研究了4种类型的蛋白质204个样品的氨基酸残基对的相关性系数,找出了可作为蛋白质结构类型特征的氨基酸残基的相关对,并用于蛋白质结构类型的预测,对于α型、β型、α/β型和α＋β型蛋白质的204个蛋白质样品的交叉测试,正确率分别为94％、89％、79％和89％,平均为88％,高于简单距离法和欧几里德距离法.

Prediction of Protein Structural Classes Based on Correlations of Amino Acid Residues

The correlation relationships among amino acid residues in protein structures are studied. Some residue pairs have larger correlation coefficients and they are not independent of protein structural classes. The correlation relationships of amino acid residues are the earmarks of protein structural classes and have special physical and chemical causes. These correlation relationships are used in the predictions of protein structures for α,β,α/β, and α＋β protein classes, in total 204 proteins. The prediction accuracy are 94％ (α-proteins),89％ (β-proteins),79％ ((α/β)-proteins) and 89％ ((α＋β)-proteins),higher than the results of the simple distance method and the Euclidean distance method.