QSAR方法预测中草药重要成分毒性的初步评价  

The evaluation of toxicity data of Chinese herbal medicine important components predicted by QSAR

作　　者：高雅[1] 姚碧云[1] 周宗灿[1] 

机构地区：[1]北京大学公共卫生学院毒理学系,食品安全毒理学研究与评价北京市重点实验室,北京100191

出　　处：《毒理学杂志》2016年第5期342-345,共4页Journal of Toxicology

基　　金：国家自然科学基金(81172693);北京市自然科学基金(7152084)

摘　　要：本研究初步评价Toxtree、T.E.S.T.、TOPKAT和Derek等QSAR平台,对中草药重要成分实验数据子库的预测可靠性。对数值型数据评价计算外部验证的相关系数Q^2,对分类型数据计算Cooper统计量。实验数据子库中,共83种成分有大鼠经口LD_(50)实验数据,计算T.E.S.T.的Q^2=0.76,TOPKAT的Q^2=0.39。共28种成分有致癌性实验数据,其中致癌性阳性16种,致癌性阴性10种,致癌性不确定2种,实验阳性物比例=16/28=57.1%。Toxtree预测准确性为57.1%,TOPKAT预测准确性为53.6%,Derek预测准确性为50.0%;共46种成分有Ames试验数据,试验阳性14种,阴性32种,实验阳性物比例14/46=30.4%,Toxtree预测准准确性为67.4%,T.E.S.T.预测准确性为76%,TOPKAT预测准确性为71.1%,Derek预测准确性为73.9%。由于不同的原因,对发育毒性、骨髓微核试验、及其他终点难以进行评价。本文提出(Q)SAR预测模型应按OECD验证5项原则进行全面的和标准化的评价。强调要定义明确毒性终点和模型应用域。毒性是复杂的生物现象,不要满足于过于简单的数值。而且,生物毒性数据均具有不确定性。To evaluate the reliability of toxicity data of Chinese herbal medicine important components predicted by Toxtree,T. E. S. T.,TOPKAT and Derek. The correlation coefficient of external validation Q^2 was calculated for evalutation of numeric data,and cooper statistics was calculated for categorical data. There were 83 important components of Chinese herbal medicine with LD_（50）（ oral rat） by experiment,the Q^2 of T. E. S. T. was 0. 76,the Q^2 of TOPKAT was 0. 39; There were 28 important components of Chinese herbal medicine with carcinogenicity by experiment,of which 16 were positive,10 were negative,2 were uncertainty,the proportion of experimental positive were 57. 1%. The prediction accuracy of Toxtree was 57. 1%,TOPKAT was 53. 6% and Derek was 50%; There were 46 important components of Chinese herbal medicine with Ames test data,of which 14 was positive,31 was negative,the proportion of experimental positive were 30. 4%,the prediction accuracy of Toxtree was 67. 4%,T. E. S. T. was 76%,TOPKAT was 71. 1% and Derek was 73. 9%. The developmental toxicity,bone marrow micronucleus test,and the other endpoints were difficult to evaluate because of different reasons.（ Q）SAR models should be evaluated standardized and comprehensively according to the five principles of OECD. It should also pay attention to define toxicity endpoints and model application domain. Toxicity is a complex biological phenomenon,couldn＇ t just be simple data.Moreover,biological toxicity data have uncertainty.

关 键 词：中草药 QSAR TOXTREE T.E.S.T. TOPKAT Derek 毒性预测 可靠性 

分 类 号：R99[医药卫生—毒理学] O6[医药卫生—药学]