基于多元线性回归的血管紧张素转化酶抑制肽定量构效关系建模研究

Modeling the QSAR of ACE Inhibitory Peptides Based on MLR

利用氨基酸结构描述符SVHEHS分别对血管紧张素转化酶(Angiotensin I-converting Enzyme,ACE)竞争性抑制二肽、三肽、四肽序列表征后,建立结构与活性的多元线性回归(MLR)模型。ACE抑制二肽模型的相关系数、交叉验证相关系数、均方根误差、外部验证相关系数分别为0.851、0.781、0.327、0.792;三肽模型分别为0.805、0.717、0.339、0.817;四肽模型分别为0.792、0.553、0.393、0.630。研究表明,运用该描述符建立的ACE抑制肽MLR模型拟合、预测能力均较好,能较好解释ACE抑制肽的活性与结构间的关系。

The structures of ACE inhibitory dipeptides,tripeptides and tetrapeptides with competitive type were characterized with amino acid descriptors SVHEHS.Furthermore,the multiple linear regression（MLR） models between the structure and the activities were built.The correlative coefficient（R2）,the crossvalidation correlative coefficient（Q2LOO）,root mean square error（RMSE）,and external validation correlative coefficient（Q2ext） were 0.851,0.781,0.327,0.792 for dipeptides model;0.805,0.717,0.339,0.817 for tripeptides model;and 0.792,0.553,0.393,0.630 for tetrapeptides model,respectively.The results showed that the MLR models constructed by this descriptor had good fitting and predictive abilities,and could be used for quantitative structure activity relationship（QSAR） research of angiotensinⅠ-converting enzyme-inhibitory peptides.