期刊文献+

FTIR对滇龙胆环烯醚萜类成分含量快速预测研究

A rapid prediction of iridoid content in Gentiana rigescens using FTIR
分享 导出
摘要 目的:建立适用于滇龙胆根、茎与叶中主要环烯醚萜类成分的含量预测模型,开发一种快速有效的质量控制方法。方法:采集云南省231份滇龙胆样品的傅里叶变换红外光谱(FTIR),用高效液相色谱(HPLC)法测定根中马钱苷酸、獐牙菜苦苷、龙胆苦苷及当药苷含量,根、茎及叶内总环烯醚萜类含量。光谱进行预处理并选取最佳预处理方式。SPXY算法将样品按2∶1分为训练集和预测集,训练集数据结合支持向量机回归(SVR)建立不同产地、部位滇龙胆环烯醚萜类含量快速预测模型,预测集数据对模型泛化能力进行外部验证。结果(:1)滇龙胆根中不同环烯醚萜类含量差异较大,其中龙胆苦苷>马钱苷酸>当药苷>獐牙菜苦苷;总体上不同部位之间总环烯醚萜含量根>叶>茎,部分产地叶中总环烯醚萜含量大于根(;2)不同产地及部位滇龙胆的红外光谱峰形及波数相似,但吸收度存在差异(;3)所有模型中,标准正态变量(SNV)+二阶导数(SD)+13点Savitzky-Golay(SG)平滑+正交信号校正(OSC)-SVR与SNV+SD+11点SG平滑+OSC-SVR对根中总环烯醚萜及龙胆苦苷含量预测效果最佳,预测集相关系数(R_p~2)分别达到0.975 8、0.971 1,RPD分别为5.56、4.99,模型可信度较高,预测值与HPLC检测值接近。结论:FTIR技术结合化学计量学对8个产地滇龙胆中环烯醚萜类成分含量进行准确预测,可作为一种快速可靠的定量方法对滇龙胆药材进行质量控制。 Objective:To establish a prediction model for the contents of major iridoids in roots,stems and leaves of Gentiana rigescens,and to develop a fast and effective quality control method.Methods:Fourier transform infrared(FTIR)spectra of 231 batches of G.rigescens from Yunnan Province were collected,and high performance liquid chromatography(HPLC)was used to determine loganic acid,swertiamarin,gentiopicroside,sweroside,and the total iridoids in roots,stems and leaves.The spectra were preprocessed,and the best pretreatment method was selected.The joint X-Y distance(SPXY)algorithm divided samples into training set and prediction set according to 2∶1.The training set was combined with support vector machine regression(SVR)to establish a rapid prediction model for total iridoids in samples from different parts and habitants.Prediction set was used for external validation of model generalization capabilities.Results:Firstly,the contents of iridoids in the roots of G.rigescens were quite different with an order of gentiopicroside>loganic acid>sweroside>swertiamarin.In general,the contents of total iridoids in different parts were in an order of roots>leaves>stems,and the contents of total iridoids in some leaves were larger than those in roots.Secondly,the infrared spectra peak shape and wavenumber of G.rigescens from different habitants and parts were similar,but the absorbance was different.Thirdly,in all models,the standard normal variable(SNV)+second derivative(SD)+13 points Savitzky-Golay(SG)smooth+orthogonal signal correction(OSC)-SVR and SNV+SD+11 points SG smooth+OSC-SVR had the best predicting results for total iridoids and gentiopicroside in the roots.The correlation coefficients(R_p~2)of the prediction set were0.975 8 and 0.971 1 with the relative percent deviations(RPD)of 5.56 and 4.99,respectively.The model showed high credibility,and the predicted values were close to the HPLC detected values.Conclusion:FTIR combined with chemometrics accurately predicted the contents of iridoids in G.rigescens from eight
作者 裴艺菲 沈涛 左智天 张庆芝 王元忠 PEI Yi-fei;SHEN Tao;ZUO Zhi-tian;ZHANG Qing-zhi;WANG Yuan-zhong(College of Traditional Chinese Medicine,Yunnan University of Chinese Medicine,Kunming 650500,China;Institute of Medicinal Plants,Yunnan Academy of Agricultural Sciences,Kunming 650200,China;College of Resources and Environment,Yuxi Normal University,Yuxi 653100,China)
出处 《药物分析杂志》 CAS CSCD 北大核心 2019年第4期615-623,共9页 Chinese Journal of Pharmaceutical Analysis
基金 国家自然科学基金(81660638,81760684) 云南省自然科学基金重点项目(2017FA049).
关键词 傅里叶变换红外光谱 支持向量机回归 滇龙胆 环烯醚萜类 马钱苷酸 獐牙菜苦苷 龙胆苦苷 当药苷 含量预测 Fourier transform infrared spectroscopy support vector regression Gentiana rigescens iridoid loganic acid swertiamarin gentiopicroside sweroside content prediction
作者简介 通信作者:张庆芝,Tel:13888396896,E-mail:ynkzqz@126.com;通信作者:王元忠,Tel:13888829994,E-mail:boletus@126.com;第一作者:裴艺菲,Tel:18404983523,E-mail:feifei950222@gmail.com.
  • 相关文献

参考文献2

二级参考文献17

共引文献2

投稿分析

相关作者

内容加载中请稍等...

相关机构

内容加载中请稍等...

相关主题

内容加载中请稍等...

浏览历史

内容加载中请稍等...
;
使用帮助 返回顶部 意见反馈