Characterization of Agarwood by Gas Chromatography–Mass Spectrometry and a Random Forest Model

Abstract

Agarwood, resin produced by wounded Aquilaria plants, is widely used in perfume and medicine. In this study, artificial agarwood induced by physical injury, chemical stimulation, fungal infection, and comprehensive stimulation, which are the common induction methods, was collected. After gas chromatography–mass spectrometry (GC–MS) analysis, the total ion chromatogram/mass ion chromatogram (TIC/MIC) and qualitative peak data from artificial agarwood were used as training sets in discriminant analysis to examine the formation modes of natural agarwood by a random forest model which was shown to be superior to discriminate the formation of natural agarwood. The results implied that natural agarwood is likely formed through four formation modes. The discriminant result from random forest model showed that 17.1% of natural agarwood samples were formed by physical injury, 22.9% by chemical stimulation, 2.9% by fungal infection, and 57.1% by comprehensive stimulation. 2-Butanone-4-phenyl-, benzenepropanoic acid, guaiol, baimuxinal, and 2-(2-phenylethyl) chromone were the main chemical markers of the agarwood. The apparent abundance from the GC–MS fingerprints was used to distinguish agarwood formed by different formation processes.

Keywords:

• Agarwood
• Aquilaria
• gas chromatography–mass spectrometry
• random forest model

Additional information

Funding

This work was supported by the Guangdong Province Science and Technology Plan Project (Grant Number 2017A020213014), by the Special Project of International Science and Technology Cooperation Guidance of Guangdong Academy of Sciences (Grant Number 2019GDASYL-0503002) and by the Open Fund Project of the State Key Laboratory of Applied Microbiology Southern China (Grant Number SKLAM002-2018).