目的 建立葛根芩连汤四味组成中药处方用量与各自煎煮液中主要药效含量的预测模型,为中药处方用量研究提供新思路.方法 以葛根芩连汤药效组分实验研究的相关文献为研究对象,纳入符合标准的文献共10篇,对数据进行整理并录入Excel表,分别采用人工神经网络回归、支持向量机回归、广义线性回归进行建模及评估,对葛根芩连汤两种配伍比例(5∶3∶3∶2和8∶3∶3∶2)下葛根素、黄芩苷、盐酸小檗碱、甘草酸含量进行预测及对比,并对煎煮过程中各药味间相互作用模拟曲线进行分析.结果 葛根素、甘草酸的神经网络回归模型与黄芩苷、盐酸小檗碱的广义线性回归模型组合最优,建立药效组分含量预测模型,模型预测及模拟曲线分析结果均显示葛根芩连汤8∶3∶3∶2用量配伍比例更为合理,更能体现组方原则,符合经方原义.结论 本模型从处方量与药效组分含量的角度为葛根芩连汤组方用量合理性验证提供了一定的客观参考依据.%Objective To establish simulation model for prescribed dosage and effective composition content of four herbs in Gegen Qinlian decoction, to provide a new way for the research of prescription dosage control of tranditional Chinese medicine(TCM).Methods Literatures associated with active components were selected as research object, there were 10 literatures were brought into this research, data were disposaled and recorded into Excel database.The model was builded and evaluated with artificial neural network regression, support vector machine regression and generalized linear regression.To forecast and compare the content of puerarin, baicalin, berberine hydrochloride and glycyrrhizic acid under two different dosage compatibility proportion(5∶3∶3∶2 and 8∶3∶3∶2), and to analyze the simulation curve of medicine interaction during the decoction.Results The optimum combination is the model of artificial neural network regression of puerarin and glycyrrhizic acid and the model of generalized linear regression of baicalin and berberine hydrochloride.Model prediction and simulation curve showed the reasonability of the 8∶3∶3∶2 proportion.Conclusion The model provide objective reference for the reasonability test of dosage of Gegen Qinlian decoction from the perspective of prescription dose and active components alignment.
展开▼
