Molecular networking (MN) is an efficient tool for natural product research. However, single MN might lead to false annotation due to the limited information, and the importance of combining MN with chromatogram is always ignored. In this study, we proposed a comprehensive MN strategy combining feature-based molecular networking (FBMN) and dual ionization mode MS/MS to improve the annotation accuracy and to achieve structural feature visualization in a chemotaxonomic chromatogram. Three steps were taken: (1) employing FBMN and dual ionization mode MS/MS to distinguish isomers and improve components' identification accuracy. (2) Using a 3-level initiative supported by in-house database to evaluate the annotation confidence. As a result, 95 compounds were successfully identified from Ginkgo biloba leaf extract (GBE) and Ginkgo biloba leaf (GBL), and 70 compounds mainly consisting of flavonoid glycosides, ginkgolides, and lignan glycosides were assigned as high-confidence molecules. (3) Building color-dependent chemotaxonomic chromatograms, to achieve component visualization by connecting FBMN with chromatogram in which the peaks of the same color indicated the compounds with similar structural features. Our research provided a new and efficient strategy for component identification and visualization of herbal medicine.
Keywords: Chromatogram; Ginkgo biloba extract; Herbal metabolites; LC-MS/MS; Molecular networking.
Copyright © 2021 Elsevier B.V. All rights reserved.