肽键的统计分析和蛋白质的二级结构

 本文对蛋白质序列的肽键进行了统计分析,计算了二肽构象参数P_α、P_β、P_c和三肽构象参数Q_α、Q_β、Q_c。在此基础上提出了由氨基酸序列预测二级结构的规则。预测的正确率达90％,优于Chou-Fasman方法。这个结果表明二肽(三肽)关联在形成蛋白质二级结构中具有明显的重要性。

STATISTICAL ANALYSIS OF PEPTIDE CORRELATION AND PREDICTION OF PROTEIN CONFORMATION

The α-helix, β-sheet and coil conformational parameters-Pα,Pβ,Pc (dipeptide) and Qα, Qβ, Qc (tripeptide) for 20 amino acids have been computed through statistical analysis of 43 protein sequences. A simple method to predict the secondary structure of protein has been suggested by utilizing these parameters. There are three prediction rules, namely, the rule of nucleation, propagation and competition. Using these rules we have succeeded in locating helical, β-sheet and coil regions with 90% of accuracy. It confirms our primary idea that the correlation of dipeptide may play an important rule in the formation of secondary structure in protein. Our method seems more successful but not more complicate than Chou-Fasman's.