摘要
建立预测类黄酮化合物抑制恶性疟原虫株活性定量的模型,并确定影响类黄酮化合物活性的主要因素。本文选用了38个结构不同的类黄酮化合物作为数据集,采用多元线性同归法及主成分分析法分析每个化合物的220个分子参数,建立最优的预测模型。比较用不同方法建立的模型,结果发现带logP参数的向后筛选法为最优方法,所建模型统计结果良好(训练集相关系数R^2=0.81,标准训练误差SEE=0.27),模型代入检验集数据时结果也令人满意(检验集相关系数R^2=0.83,标准检验误差SEP=0.39),可靠性和预测性较强。脂水分配系数的对数logP为模型重要影响参数。建模和确定影响因素有助于筛选新型类黄酮抗疟疾药物和研发。
To build a predictable mathematic model of inhibition of Plasmodiumfalciparum by flavonoids and determine the key factor of inhibition of flavonoids,we built a dataset composed of 38 flavonoids with diversiform structures,then regressed the 220 molecular indices by multivariate linear regression and principal component analysis methods and finally got the best predictable mathematic models of their own.Backward regression analysis was found to be the optimal regression method compared with other multivariate linear regressions. That Enter regression model demonstrated satisfactory statistical results(R^2=0.81,SEE=0.27),whose proper predictability was validated by the independent test set(R^2=0.83,SEP=0.39).The octanol-water partition coefficients logP is an important molecular indices.The key influencing factors were identified,which were valuable and helpful for further researching and developing new anti-malaria drug of flavonoids.
出处
《计算机与应用化学》
CAS
CSCD
北大核心
2010年第4期485-491,共7页
Computers and Applied Chemistry
基金
大连理工大学青年教师培养基金资助课题(No:1000-893231)
大连理工大学博士科研启动基金资助课题(No:1000-893361)
自然科学基金(No:10801025)
关键词
类黄酮
恶性疟原虫
QSAR
多元线性回归
主成分分析
flavonoids
Plasmodium falciparum
QSAR
multivariate linear regression
principal component analysis