引用本文: |
-
王青艳,谢能中,许晓东.药物分子设计中定量结构-活性关系计算方法的研究[J].广西科学,2014,21(1):6-11. [点击复制]
- WANG Qing-yan,XIE Neng-zhong,XU Xiao-dong.Study of Mathematical Method in the Quantitative Structure-activity Relationship for Drug Design[J].Guangxi Sciences,2014,21(1):6-11. [点击复制]
|
|
摘要: |
[目的]发展一种新的定量结构-活性关系(简称"构效关系"QSAR)计算方法。[方法]双层QSAR预测模型的第1层是分子的理化性质参数,第2层是分子片段结构参数。两组权重系数{ak}和{bl}分别指派给理化参数和片段参数。在双层QSAR预测模型中引入迭代的双最小二乘(IDLS)计算方法,把单层次的构效关系发展成二层次、双方向的预测网络;在训练集中两组系数用最小二乘法(或偏最小二乘法、主成分分析等)交替求解,直至收敛。[结论]二层次、双方向的QSAR预测模型不仅能够提高预测能力,具有一定的信息反馈和学习功能,而且赋予QSAR更多的信息分析能力。 |
关键词: 药物设计 构效关系 分子片段 理化性质 |
DOI: |
投稿时间:2013-09-04修订日期:2013-10-11 |
基金项目:国家自然科学基金项目(项目代号:31370716和31360207)资助。 |
|
Study of Mathematical Method in the Quantitative Structure-activity Relationship for Drug Design |
WANG Qing-yan, XIE Neng-zhong, XU Xiao-dong
|
(Guangxi Academy of Sciences, State Key Laboratory of Non-Food Biomass and Enzyme Technology, National Engineering Research Center for Non-food Biorefinery, Guangxi Biomass Industrialization Engineering Institute, Guangxi Key Laboratory of Biorefinery, Nanning, Guangxi, 530007, China) |
Abstract: |
[Objective] Quantitative structure-activity relationship (QSAR) is a widely used approach in chemistry, biology, and environment science.In recent years the iterative double least square (IDLS) method was applied in the QSAR by authors.[Method] The simple and single direction prediction method was reformed to a two-level and double direction prediction net work.In the two-level QSAR model one level is physicochemical parameters {ak}, and the other is the molecular fragments {bl}.In a training set the two coefficient sets are solved by using least square (or partial least square and principal component analysis) iteratively.[Conclusion] The two-level and two-direction QSAR method not only enhances the prediction power, but also endows more information analysis ability to QSAR. |
Key words: drug design QSAR molecular fragments physicochemical properties |