Utilizing High-Resolution Mass Spectrometry Data Mining Strategy on R Programming Language for Rapid Annotation of Absorbed Prototypes and Metabolites of Gypenosides

Abstract

Rapid and accurate annotation of the complex compounds and metabolites of natural products is still a major challenge. In this study, based on the established Gypenosides component database, an intelligent strategy was developed by integrating R programming language, sample selection, virtual metabolite library, polygon mass deficit filtration (PMDF) and Kendrick mass defect filtering (KMDF) for rapid and accurate characterization of metabolites in natural products. In addition, the established annotation strategy was applied to annotate the prototypes and metabolites of the serum of Gypenosides-administrated mice. The feasibility of the proposed strategy was verified by taking Gypenoside ⅬⅩⅩⅤ as an example. Results: 719 potential prototypes and metabolites of Gypenosides were rapidly screened from mice serum by PMDF. And 38 prototype components and 108 metabolites in mice serum was accurately annotated by using the self-built virtual metabolite library and KMDF. The metabolic pathway of Gypenoside ⅬⅩⅩⅤ in vivo was analyzed by the integrated strategy, and the prototype and 8 metabolites of Gypenoside ⅬⅩⅩⅤ were further confirmed, indicating that the established annotation strategy could be applied to annotate the prototype components and metabolites of Gypenosides and the proposed strategy is available. Significance: This study provides an available strategy for rapid and accurate identification of prototypes and metabolites of natural products, and elucidates the metabolic pathways of GPs in vivo for the first time.

Keywords: R programming language, Virtual metabolite library, Polygonal mass defect, Kendrick mass defect, Gypenosides, Metabolite annotation