摘要
用量子化学从头算RHF方法在6-31G(d)水平下,对吡啶类分子进行构型全优化,用优化得到的量化参数构建回归模型,预测吡啶类化合物的PKa值．实验结果表明用其模型预测的相关系数为0．9548、最大绝对误差为1．139、最小绝对误差为0．03、误差平方和为12．671．实验证明所构造的回归模型在预测吡啶类化合物的PKa值中得到满意的结果．
The quantitifcational parameters which obtained from the ab initio calculation to Pyridines Compounds by RHF／6-31G method in G98w software package are used as the structural descriptors for Pyridines Compounds. A multiple linear quantitative equation between the pica and the structural descriptors of Pyridines Compounds has been established with multiple linear regression method. The multiple regression coefficient is 0.9548, the most of the absolute error is 1.139, the smallest of the absolute error is 0.03, and the sum of the double error is 12.671. The multiple regression model is proved to be satisfied to predicting the PiCa of Pyridines Compounds by compared with experimental data.
出处
《青海师专学报》
2006年第5期77-80,共4页
Journal of Qinghai Junior Teachers' College
关键词
吡啶类化合物
量化化学计算
PKA
多元线性回归
Pyridines Compounds
Quantum Chemical Calculation
PlEa
multiple linear regression
