Chemometric discrimination of the geographical origin of licorice in China by untargeted metabolomics

Licorice is famous as a herbal medicine and food sweetener. This study provided a comprehensive strategy for investigating the quality of licorice based on untargeted metabolomics. A new strategy for identifying metabolite was developed, including fragment ion identification algorithm and ion fusion algorithm. The results showed that it can accurately integrate mass spectra from positive and negative ion modes to benefit metabolite identification. Based on the strategy, a number of significant difference metabolites were identified among licorice samples and 9 metabolites were confirmed by standards. Additionally, the geographical discrimination models of licorice samples were comprehensively investigated by chemometric methods. The results indicated that the supporting vector machine provided the best performance, with a prediction accuracy above 80%. The study results supported the conclusion that the quality of licorice from different regions in China was inconsistent.