摘要
利用HyperChem软件包中的PM3半经验量子化学计算方法,将50个细胞色素氧化酶P450(CPY1A2)抑制剂,作几何结构优化和计算,选取一些可能会影响及pIC_(50)值的22个物理化学和电子结构参数,作为描述分子的变量进行QSAR研究,利用偏最小二乘(PLS),穷举回归(QLS)及人工神经网络方法(CGANN),以建立物理化学和电子结构等参数与其抑制活性之间的QSAR模型。结果表明:该类化合物的折射率较大、生成热较小,且log P取值在2．1～2．9之间时,具有较高的抑制细胞色素P450 CYP1A2的活性。
50 cytochrome P450 1 A2 inhibitors have been optimized and calculated using the empirical quantum chemistry method PM3 in the software bag of HyperChem. QSAR studies has been approached based on 22 parameters including physical-chemical and electronic structure parameters which may work on the pIC50 values, using PLS, QLS and ANN methods. Among the 20 QSAR equations with the least root mean square deviation, the appearance probabilities of parameters log P, （ log P） 2, refractivity, and heat of forma- tion were highest. The results show that it may be quadratic functional relationship between log P and the activities of cytochrome P450 1A2 inhibitors. When the inhibitors have a larger refractivity and heat of formation, and the value of log P in the span 2. 1 to 2. 9, the activities of cytochrome P450 1 A2 inhibitors will be higher.
出处
《计算机与应用化学》
CAS
CSCD
北大核心
2007年第2期177-181,共5页
Computers and Applied Chemistry
基金
河南省杰出青年科学基金支助项目(061200260)
化学生物传感与计量学国家重点实验室(湖南大学)开放课题
河南省高校青年骨干教师资助计划
