摘要
选取已测活性的17个穿心莲内酯衍生物,利用高斯软件计算得到其结构参数,通过主成分分析筛选出主要影响参数,进一步利用BP神经网络对建立的活性评价模型进行验证.建立了穿心莲内酯衍生物的构效关系,为进一步开发具有更高活性的穿心莲内酯类药物提供理论依据.
The confirmed activity data of 17 andrographolide derivatives were selected. Structural param- eters were calculated by Gaussian software. According to the principal component analysis, the main pa- rameters were chosen. Then the proposed activity evaluation model was verified through a BP neural net- work. This QSAR modeling provided the theoretical basis for better andrographolide drugs synthesis.
出处
《郑州大学学报(理学版)》
CAS
北大核心
2015年第1期103-106,共4页
Journal of Zhengzhou University:Natural Science Edition
基金
国家自然科学基金资助项目
编号J1210060