Assessing the Quality of Calyx of Physalis alkekengi L. var. franchetii Based on Quantitative Analysis of Q-Marker Combined with Chemometrics and Machine Learning Algorithms

Meiqi Liu,Lizhi Wang,Ziying Qiu,Yanru Deng,Xiaoliang Ren,Xiaoran Zhao,Lili Sun,Isabel Mafra

doi:10.1155/2021/8502929

Abstract

Physalis alkekengi L. var. franchetii (PALF) is a traditional Chinese medicine, which is well known for its antimicrobial, anti-inflammatory, antipyretic, and expectorant properties. Its fruits and fruiting calyxes are used as dietary supplements and traditional herbs in China. However, the quality of calyxes is uneven, and it is prone to getting moldy or moth-eaten during storage. High-performance liquid chromatography (HPLC) fingerprints and multivariate chemometric methods were combined to evaluate quality, and three representative compounds were chosen as the quality markers (Q-markers). Hierarchical cluster analysis (HCA) and principal component analysis (PCA) provided a clear discrimination of PALF samples. Through further verification by partial least squares discriminant analysis (PLS-DA), backpropagation artificial neural network (BP-ANN), machine learning, and combination with the determination of the content, biology, and pharmacology effect judgment, galuteolin, rutin, and physalin O could be used as Q-markers that their contents affect the quality of PALF grade evaluation. A simple method was established to rapidly assess the quality of PALF that is important for its clinical application and storage and provide a reference for evaluating the quality of materials used in Chinese medicine.

Highlights

Physalis alkekengi L. var. franchetii (PALF) consists of dried persistent calyx, or persistent calyx of the fruit of Physalis alkekengi L. var. franchetii (Mast.) Makino (Solanaceae)
Method as shown in Figure 1. e 31 common peaks in the different samples were obtained by multiple points and automatically matched, and the similarity was characterized by the correlation coefficient
High-performance liquid chromatography (HPLC) workstation data management software was used to obtain the peak area, retention time, and other related information of PALF. e obtained data matrix was imported into SIMCA-P14.0. e values of R2X and Q2 of the principal component analysis (PCA) model were 0.784 and 0.908, respectively. e PCA score chart (Figure 2(a)) shows that the samples were divided into two categories. e G1 samples were characterized by orange or orange-red in the permanent calyxes and almost no insects or mold spots

Summary

Introduction

Physalis alkekengi L. var. franchetii (PALF) consists of dried persistent calyx, or persistent calyx of the fruit of Physalis alkekengi L. var. franchetii (Mast.) Makino (Solanaceae). Franchetii (PALF) consists of dried persistent calyx, or persistent calyx of the fruit of Physalis alkekengi L. var. Its calyxes in particular are considered components of some important medicines Tea made from it can relieve cough and reduce phlegm as well as detoxifying the body [2]. It consists of physalins and flavonoids as well as terpenoids, alkaloids, aliphatic compounds, sterols, phenylpropanoids, and phenolic acids. Such physalins, as physalin A, have a significant inhibitory effect on the proliferation of a variety of tumors [3,4,5,6]. PALF has antimicrobial, anti-inflammatory, antioxidation, and hypoglycemic effects [7,8,9,10,11]

Methods

Results

Conclusion