Abstract: In order to study the effect of fermentation by Lactobacillus paracasei NXU-19004 on the quality characteristics of goji juice, non-targeted metabolomics based on gas chromatography-mass spectrometry (GC-MS) was used to identify and analyze the differential metabolites of fermented goji juice samples at four different fermentation times, and principal component analysis (PCA) and partial least squares-discriminant analysis (PLS-DA) were used to distinguish the four samples. Then the metabolic pathways of the differential metabolites were annotated in the KEGG database to identify the key metabolites. The results showed that a total of 63 differential non-volatile metabolites were detected in the four samples. Among them, the average relative abundance of amino acids, organic acids and lipids in fermented goji juice showed an overall increasing trend with fermentation time. The results of multivariate statistical analysis showed that there were significant differences between the samples, and an obvious clustering area was observed for each sample, indicating that the samples can be discriminated based on the differential metabolites. A total of 47 metabolic pathways were identified, and 9 key pathways through which more differential metabolites could be enriched were observed, most of which were involved in amino acid metabolism. It could be concluded that fermentation time had the greatest influence on the metabolism of amino acids in fermented goji juice. It was further confirmed that the key metabolites were alanine, limonene, benzaldehyde, benzoic acid, nicotinamide, niacin, squalene, proline, sarcosine and 2-phenylacetamide. The results of this study provide a theoretical basis for further studies on the factors affecting the formation of differential metabolites in goji juice during lactic acid bacteria fermentation.

Key words: goji juice; Lactobacillus paracasei; non-targeted metabolomics; metabolic pathways; key metabolites