摘要 目的采用球面对称设计及遗传神经网络优选黄芪-红花药对的水提工艺。方法将黄芪-红花药对水提液中8种有效成分(羟基红花黄色素A、紫丁香苷、毛蕊异黄酮苷、脱水红花黄色素B、山柰酚-3-O-芸香糖苷、芒柄花苷、毛蕊异黄酮、黄芪甲苷)的含量进行赋权,以计算得到的综合得分作为水提工艺的评价指标。在单因素实验基础上,选择液料比、提取时间、提取温度3个因素,依据球面对称设计进行实验,并结合遗传神经网络模型进一步优选黄芪-红花药对水提工艺参数。结果球面对称设计得到的最佳提取条件为液料比20∶1、提取时间30 min、提取温度87℃、提取次数2次,其综合评分为4.877;以遗传神经网络得到的最佳提取条件为液料比22.7∶1、提取时间30 min、提取温度87℃、提取次数2次,经验证该条件下的综合得分为5.004。实验结果显示,遗传神经网络能够提高水提工艺条件的综合评分。结论球面对称设计和遗传神经网络能够优化黄芪-红花药对水提工艺参数,可为中药及其制剂中多种有效成分提取的工艺优选提供参考。 Objective To determine the appropriate aqueous extract craft for Astragali Radix-Carthami Flos via spherical symmetry design and genetic neural network.Methods In this study,the contents of eight active components,including hydroxysafflor yellow A,syringin,calycosin-7-glycoside,anhydrosafflor yellow B,kaempferol-3-O-rutinoside,ononin,calycosin,and astragaloside IV,in the aqueous extract of Astragali Radix-Carthami Flos were weighted,and the calculated comprehensive score was applied as the evaluation index.According to the single factor experiment,the ratio of material to liquid,extraction duration and extraction temperature were taken as three factors of spherical symmetric design for the experiment.The genetic neural network model was established to further optimize the aqueous extraction process parameters of Astragali Radix-Carthami Flos.Results The optimal extraction conditions obtained by spherical symmetry design were liquid to solid ratio of 20:1,extracting at 87℃,decocting twice and 30 min every time,and the comprehensive score was 4.877;The optimum extraction conditions obtained by the genetic neural network were liquid to solid ratio of 22.7:1,extracting at 87℃,decocting twice and 30 min every time,and the comprehensive score was 5.004.The experimental results showed that the genetic neural network can improve the comprehensive score of water extraction process conditions.Conclusion Spherical symmetric design combined with genetic neural network mathematical model has ability to optimize the water extraction process parameters of Astragali Radix-Carthami Flos,which provides a reference for optimizing the extraction process of various effective components from traditional Chinese medicine and its preparations.
机构地区 浙江中医药大学
出处 《中草药》 CAS CSCD 北大核心 2021年第8期2257-2266,共10页 Chinese Traditional and Herbal Drugs
基金 国家自然科学基金项目(81873226) 国家科技重大专项重大新药创制(2019ZX09301101) 浙江省自然科学基金项目(LZ18H270001) 浙江省“万人计划”科技创新领军人才项目(2019)。
关键词 黄芪-红花药对 球面对称设计 遗传神经网络 工艺优化 羟基红花黄色素A 紫丁香苷 毛蕊异黄酮苷 脱水红花黄色素B 山柰酚-3-O-芸香糖苷 芒柄花苷 毛蕊异黄酮 黄芪甲苷 Astragali Radix-Carthami Flos drug pair spherical symmetry design genetic neural network process optimization hydroxysafflor yellow A syringin calycosin-7-glycoside anhydrosafflor yellow B kaempferol-3-O-rutinoside ononin calycosin astragaloside IV
