Objective:This paper discusses the composition of prescription qualitative,quantitative design principles and methods based on herbal property combination,describing the method application in new prescription design.M...Objective:This paper discusses the composition of prescription qualitative,quantitative design principles and methods based on herbal property combination,describing the method application in new prescription design.Method:Qualitative property-combination pattern(PP)calculation was based on bipartite graphing and performing a greedy algorithm which was designed to optimize obtaining a new herbal prescription.Quantitative PP calculation was based on the qualitative computation.To calculate the Euclidean distance for the PP of the new prescription,an optimized algorithm for solving the unknown minimum Euclidean distance was used with,the new weighted proportions.Finally,non-linear optimization software was used to find the minimum Euclidean distance.Results:Using the PP of classic prescription Large Yin-Nourishing Pill,applying quantitative PP calculation a new prescription was created.Mathematical algorithms based on property combinations of traditional Chinese herbs can be applied to identify compatibility and synergies of herbs within prescriptions,especially classic formulas.Conclusion:In silico methods can then be used to create new prescriptions or modify existing ones depending on need.This type of automated approach may increase efficiency in designing new drugs based on Chinese herbs.展开更多
目的将神经网络模型与传统中药药性理论(四性、五味、归经)相结合来预测分析中药肾毒性,方法通过文献检索具有肾毒性证据的中药,并将《中华本草(精选本)》中去除上述肾毒性中药的其他中药作为非肾毒性中药纳入数据。以《中华本草》为标...目的将神经网络模型与传统中药药性理论(四性、五味、归经)相结合来预测分析中药肾毒性,方法通过文献检索具有肾毒性证据的中药,并将《中华本草(精选本)》中去除上述肾毒性中药的其他中药作为非肾毒性中药纳入数据。以《中华本草》为标准,确定每味中药的四性、五味和归经归属,分别进行肾毒性/非肾毒性中药与其四性、五味、归经因素的相关性检验,筛选出相关性变量因素,用于构建神经网络模型(Neural Networks Model,NNM)。同时,绘制模型的"受试者工作特征曲线"(Receiver Operator Characteristic Curve,ROC曲线),并计算曲线下面积(Area Under the Curve,AUC),用于评估模型的预测能力。结果肾毒性/非肾毒性中药与四性、五味归属具有相关性(P<0.05),与归经归属无相关性(P>0.05)。NNM结果显示,热性、辛味、温性和苦味是影响中药肾毒性的前4位重要因素,热性排在重要性第1位,模型ROC曲线的AUC计算结果为0.739。结论将传统中药理论与现代数理统计方法相结合建立的中药肾毒性神经网络模型具有一定的预测性,该建模方法可为中药肾毒性及中药毒理学研究提供一定的参考。展开更多
基金The authors gratefully acknowledge the support of this work by the National Natural Science Foundation of China(No.81430094,81173568,81373985).
文摘Objective:This paper discusses the composition of prescription qualitative,quantitative design principles and methods based on herbal property combination,describing the method application in new prescription design.Method:Qualitative property-combination pattern(PP)calculation was based on bipartite graphing and performing a greedy algorithm which was designed to optimize obtaining a new herbal prescription.Quantitative PP calculation was based on the qualitative computation.To calculate the Euclidean distance for the PP of the new prescription,an optimized algorithm for solving the unknown minimum Euclidean distance was used with,the new weighted proportions.Finally,non-linear optimization software was used to find the minimum Euclidean distance.Results:Using the PP of classic prescription Large Yin-Nourishing Pill,applying quantitative PP calculation a new prescription was created.Mathematical algorithms based on property combinations of traditional Chinese herbs can be applied to identify compatibility and synergies of herbs within prescriptions,especially classic formulas.Conclusion:In silico methods can then be used to create new prescriptions or modify existing ones depending on need.This type of automated approach may increase efficiency in designing new drugs based on Chinese herbs.
文摘目的将神经网络模型与传统中药药性理论(四性、五味、归经)相结合来预测分析中药肾毒性,方法通过文献检索具有肾毒性证据的中药,并将《中华本草(精选本)》中去除上述肾毒性中药的其他中药作为非肾毒性中药纳入数据。以《中华本草》为标准,确定每味中药的四性、五味和归经归属,分别进行肾毒性/非肾毒性中药与其四性、五味、归经因素的相关性检验,筛选出相关性变量因素,用于构建神经网络模型(Neural Networks Model,NNM)。同时,绘制模型的"受试者工作特征曲线"(Receiver Operator Characteristic Curve,ROC曲线),并计算曲线下面积(Area Under the Curve,AUC),用于评估模型的预测能力。结果肾毒性/非肾毒性中药与四性、五味归属具有相关性(P<0.05),与归经归属无相关性(P>0.05)。NNM结果显示,热性、辛味、温性和苦味是影响中药肾毒性的前4位重要因素,热性排在重要性第1位,模型ROC曲线的AUC计算结果为0.739。结论将传统中药理论与现代数理统计方法相结合建立的中药肾毒性神经网络模型具有一定的预测性,该建模方法可为中药肾毒性及中药毒理学研究提供一定的参考。