化学计量学
线性判别分析
主成分分析
偏最小二乘回归
化学
模式识别(心理学)
衰减全反射
色谱法
气相色谱-质谱法
传统医学
质谱法
人工智能
数学
红外光谱学
计算机科学
统计
医学
有机化学
作者
Le Wang,Xiang Li,Yu Wang,Xueyang Ren,Xiaoyun Liu,Ying Dong,Jiamu Ma,Ruolan Song,Jing Wei,Axiang Yu,Qiqi Fan,Dongjie Shan,Jing Yao,Gaimei She
标识
DOI:10.1016/j.jep.2021.114422
摘要
Curcumae Radix (Yujin) has a long medicinal use history in China, which is used to cure diseases like jaundice, cholelithiasis caused by dampness-heat of gallbladder and liver, and so on. It comes from the dried tuberous roots of C. kwangsiensis (Guiyujin), C. longa (Huangyujin), C. phaeocaulis (Lvyujin) and C. wenyujin (Wenyujin). Though there are differences in chemical compositions and pharmacological activities among the four species of Yujin, they have not been differentiated well in clinical application due to their similar morphological characterizations.In this study, the four species of Yujin were rapidly and accurately discriminated. The potential volatile markers for varietal recognition were identified.Attenuated total reflection fourier transformed infrared (ATR-FTIR) spectroscopy combined with chemometrics was used to rapidly discriminate the four species of Yujin. Headspace-gas chromatography-mass spectrometry (HS-GC-MS) technology coupled with chemometrics was employed to characterize volatile profiling, differentiate species and select potential markers for varietal recognition of Yujin.By applying PCA (principal components analysis) and HCA (hierarchical cluster analysis), HS-GC-MS realized complete differentiation of the four species of Yujin, while ATR-FTIR only recognized Guiyuijin. Back propagation neural network (BP-NN), KNN (K-nearest neighbor) and LDA (linear discriminant analysis) models based on spectral data achieved 100% discriminant accuracies. Support vector machines (SVM), KNN and PLS-DA (partial least square discriminant analysis) models based on volatile compounds also realized 100% discriminant accuracies. Additionally, the potential volatile markers for varietal recognition of Yujin were screened using PLS-DA, including 2 for Guiyujin, 6 for Lvyujin, 9 for Wenyujin and 13 for Huangyujin.The present study developed reliable methods for the varietal discrimination and volatile compounds characterization of Yujin, which will provide references for its quality control and clinical efficacy.
科研通智能强力驱动
Strongly Powered by AbleSci AI