Comprehensive study of different grades of Lonicerae japonicae flos: GC-IMS, UPLC-HRMS, and machine learning

Abstract

Lonicerae japonicae flos (LJF), a traditional Chinese medicine with the homology characteristics of food and medicine, is commonly classified into three grades by sensory evaluation. Currently, few studies focus on the difference in their chemical compositions. This study aims to analyze total polyphenols and flavonoids in different grades of LJF, and determine their antioxidant activity. It can be observed that higher grades of LJF have higher content of main components and stronger antioxidant activity. The volatile and non-volatile compounds are detected using HS-GC-IMS and UPLCHRMS, a total of 12 differential volatile compounds and 15 differential non-volatile compounds are screened. Finally, eleven machine learning models are established to quickly identify LJF grades, and the results show that SVM-P, LDA, and SVM-L have potential applications for the identification of LJF grades. This study provides a detailed metabolite profile of LJF grades and establishes an identification method for grade classification.

Keywords

• Lonicera japonica Thunb
• GC-IMS
• LC-MS
• machine learning
• different grades