摘要
从中药的整体性出发,开展中药拉曼谱图与其寒热药性的相关性分析,并对中药寒热药性进行统计判别研究。本研究选取寒凉性中药109种、温热性中药128种,共计237种中药;经样品前处理后,利用如海光电SEED 3000近红外拉曼光谱仪进行检测,得到每味中药的拉曼谱图;并对量化后的中药拉曼数据进行特征筛选和统计检验,筛选出与寒热药性密切相关的特征拉曼位移及其峰强,然后基于5种算法进行寒热药性的判别建模。经比较分析发现,相较于其他模型,随机森林(RF)模型展现出最佳的效果,对测试集判别的正确率高于90%,曲线下面积(AUC)和精确度大于0.90。本研究基于大样本量中药的分析,中药的拉曼数据与其寒热药性之间具有显著的相关性,可作为药性表征指标,结合RF算法进行寒热药性的判别分析。
From the holistic view of traditional Chinese medicine(TCM),this study carried out a correlation analysis between the Raman spectra of TCM drugs and cold-heat medicinal properties and conducted a statistical discrimination analysis of the cold-heat medicinal properties of TCM drugs.A total of 237 TCM drugs were selected, including 109 cold and 128 heat ones.After sample pre-treatment, Raman spectra of TCM drugs were obtained by using SEED 3000 near-infrared Raman spectrometer(Oceanhood, China),and the characteristic Raman shifts and peak intensities closely related to the cold-heat medicinal properties were screened out by the characteristics screening and statistical test of the quantified Raman data of TCM drugs.The discriminant modeling of cold-heat medicinal properties was carried out based on five algorithms.Compared with other models, the random forest(RF) model showed the optimal results, with the correct rate of the test set higher than 90% and an AUC and precision greater than 0.90.Based on the analysis of a large sample size of TCM drugs, the findings of this study suggest that there is a significant correlation between the Raman data of TCM drugs and their cold-heat medicinal properties, which can be used as the characterization index of medicinal properties and facilitate the discriminant analysis of cold-heat medicinal properties combined with RF algorithm.
作者
梁浩
纪徐维晟
王献瑞
王耘
LIANG Hao;JI Xuweisheng;WANG Xianrui;WANG Yun(Research Center of TCM-Information Engineering,School of Chinese Materia Medica,Beijing University of Chinese Medicine,Beijing 102488,China;School of Life Sciences,Beijing University of Chinese Medicine,Beijing 102488,China)
出处
《世界中医药》
CAS
2023年第1期65-69,共5页
World Chinese Medicine
基金
国家自然科学基金项目(81973495)——基于吸引子的中药有效成分群辨识与作用机理解析方法研究。
关键词
中药
拉曼光谱
药性
寒热
随机森林
判别分析
特征筛选
TCM drugs
Raman spectroscopy
Medicinal properties
Cold-heat
Random forest
Discriminant analysis
Characteristics screening
