Integrated metabolomics and high-throughput sequencing to explore the dynamic correlations between flavor related metabolites and bacterial succession in the process of Mongolian cheese production.

Untargeted metabolomics was used to investigate the changes in small molecules of non-volatile metabolites and related mechanisms in the production process of Mongolian cheese. In addition, the changes in volatile compounds, free amino acids and bacterial community succession were also studied. A total of 37 differential metabolites were screened at 5 different stages of the production process by multivariate analysis of metabolomics datasets. The differential metabolites associated with amino acids, glycerophospholipid and pyrimidine metabolism were found to be mainly changed. In addition, the total nine functional core bacterial genera that significantly influenced compounds distribution in the cheese production process were identified based on correlation analysis and bidirectional orthogonal partial least squares (O2PLS) analysis. Lactococcus and Acinetobacter had positive effects on the formation of flavor volatiles and beneficial amino acids (p < 0.05); Streptococcus and Serratia showed strong correlations with amino acid catabolism (p < 0.05). This research revealed comprehensive insights into the metabolite alterations, bacterial succession and their dynamic correlations in the cheese process and would provide a theoretical basis for better quality control of Mongolian cheese production.