目的 使用数据处理及分析的相关技术对牛黄类药材的红外光谱数据进行鉴别研究,鉴别牛黄的天然品与其各种替代品。方法 选取一定的红外光谱数据范围,对数据进行一阶导数、标准正态变量校正以及自动缩放等预处理,然后对处理过后的数据分别进行主成分分析、层次聚类分析、K最近邻分析、偏最小二乘法判别分析和支持向量机判别分析。结果 主成分分析和层次聚类分析都显示样品的数据具有良好的特征性,聚类情况良好。通过各批样品红外光谱数据的主成分分析变量载荷图,结合牛黄类药材中化学成分结构与红外光谱吸收谱段的关系对各种牛黄类药材红外光谱特征谱段进行了分析和讨论。K最近邻分析、偏最小二乘法判别分析和支持向量机判别分析对验证集的预测结果均100%准确。结论 红外光谱具有良好的特征性,能够实现对牛黄类药材的准确鉴别。
Abstract
OBJECTIVE Bovis Calculus is a valuable Chinese medicine in China. The identification of Bovis Calculus and its substitutes is one ofthe difficulties in the research of traditional Chinese medicine. In view of this, we aim to establish and evaluate the identification methods of succession medicinal substances of Bovis Calculus with infrared spectrum information data based on techniques of data handling and data analysis. METHODS Two ranges of infrared spectra data were selected and preprocessed by first derivative, standard normal variatetransformation(SNV) and auto-scale before analysis. In the analytical process, principal component analysis(PCA), hierarchical cluster analysis(HCA),K-nearest neighbors(KNN), partial least squares-discrimination analysis(PLS-DA), and support vector machines-discrimination analysis(SVM-DA) were used.Based on the loading plot in PCA and the relationship between the structure of chemical composition and the infrared spectrum absorption band, the infrared spectrum characteristic bands of various Bovis Calculus were analyzed and discussed.RESULTS PCA and HCA indicated that the infrared spectra of samples had good characteristics for identification and the cluster of samples was correct. As for KNN, PLS-DA, and SVM-DA, the prediction accuracies of test setall were 100%.CONCLUSION Infrared spectrum has good characteristics, which can realize the accurate identification of succession medicinal substances of Bovis Calculus.
关键词
牛黄 /
红外光谱 /
数据处理 /
数据分析 /
特征性
{{custom_keyword}} /
Key words
Bovis Calculus /
infrared spectrum /
data handling /
data analysis /
characteristic
{{custom_keyword}} /
中图分类号：
R284
{{custom_clc.code}}
({{custom_clc.text}})
{{custom_sec.title}}
{{custom_sec.title}}
{{custom_sec.content}}
参考文献
[1] Ch. P(2020) Vol I(中国药典2020年版. 一部). 2020: 72-73
[2] ZOU Q W, SHI Y, LIU W, et al. The research of quantity test method of various components in succession medicinal substances of cow-bezoar . Chin J Pharm Anal(药物分析杂志),2015,35(1):8-15
[3] ZHANG Q M,LI D Y,YAN K D,et al. Comparative study of the chemical compositon of cultured Bovis Calculus and natural bezoar Ⅱ. The comparison of lamelarqualitation of unconjugated and conjugated bile acid . J Biochem Pharm(生化药物杂志), 1991(2):50-52
[4] ZHANG Q M,YAN K D,TIAN S J,et al. Comparative study of the chemical compositon of cultured Calculus Bovis and natural bezoar Ⅳ. The determination and comparison of microelement . J Chin Med Mater(中药材), 1991,14(5):15-17
[5] ZHANG Q M,YAN K D,TIAN S J,et al. Comparative study of the chemical compositon of cultured Calculus Bovis and natural bezoar Ⅴ. The determination and comparison of unconjugated and total amino acid . J Chin Med Mater(中药材),1991,14(9):15-17
[6] SHI Y, SUN D M, WEI F, et al. Quantification of main bileacids in in vitro cultivated Calculus bovis by high-performance liquid chromatography with pre-column derivatization . Chin J Pharm Anal(药物分析杂志),2016,36(11):2046-2050
[7] XIONG J, ZHENG T J, SHI Y, et al. Analysis of the fingerprint profile of bioactive constituents of traditional Chinese medicinal materials derived from animal bile using the HPLC-ELSD and chemometric methods: An application of a reference scaleplate. J Pharm Biomed Anal, 2019, 174:50-56
[8] LI Y, WU R R, YU B H, et al. A review on applications of infrared spectroscopy to the study of Traditional Chinese Medicine. Spectrosc Spectr Anal(光谱学与光谱分析), 2006, 26(10):1846-1849
[9] YUAN D, JIN F, YU S C, et al. FI-IR spectroscopy in identification of bear biles and other animal biles . Mod Chin Med(中国现代中药),2010,12(1):30-32
[10] OTTO M. Chemometrics: Statistics and Computer Application in Analytical Chemical(Third Edition) . Weinheim: Wiley-VCH Press, 1999:140-208
[11] BISHOP M C. Pattern Recognition and Machine Learning. New York: Springer,2006:1-3
[12] GRON AURLIEN. Hands-On Machine Learning with Scikit-Learn, Keras, and Tensor Flow(Second Edition). Sebastopol: O′Reilly Press, 2019:237-274
[13] ZHOU Q W, SHI Y, WEI F, et al. Study on chemical compositionand pharmacological effects of Bovis Calculus medicinal material . Chin Pharm Aff(中国药事),2014,28(6):646-650.
[14] XIONGJ, HE L, WEI F,et al. Analysis and research of succession medicinal substances of cow-bezoar and cattle bile with infrared spectroscopy . Chin J Pharm Anal(药物分析杂志), 2021, 41(4):735-740.
[15] SUN X M, PEI D, HUANG X Y, et al. Variable selection and chemometric models for discriminating symptomatic gout based on a metabolic target analysis . J Chemom,2018,32(5):DOI: 10.1002/cem.2984
[16] KANTARDZIC M. Data Mining: Concepts, Models, Methods, and Algorithms(Third Edition). Piscataway: IEEE Press, 2020:1-9.
[17] ISSAQ H J, WAYBRIGHT T J, MUSCHIK G M, et al. Analytical and statistical approaches to metabolomics research. J Sep Sci, 2009,32:2183-2199.
{{custom_fnGroup.title_cn}}
脚注
{{custom_fn.content}}
基金
2019年度国家药典委员会标准制修订研究课题资助(2019Z059)
{{custom_fund}}
PDF(3494 KB)
