Abstract: Untargeted metabolomics and genomics approaches were integrated to systematically elucidate the accumulation patterns and potential microbial origins of non-volatile beneficial compounds during the acetic acid fermentation of Shanxi aged vinegar. Metabolomic analysis identified a total of 2 088 compounds, among which 1 247 were significantly differential compounds. These encompassed various flavor precursors and bioactive components with medicinal value, exhibiting dynamic accumulation patterns throughout the fermentation process. Notably, on day 9, Cupei was enriched with 49 bio-functional compounds and 8 key flavor compounds. Correlation analysis further revealed that Acetilactobacillus jinshanensis, Acetobacter pomorum, Acetobacter sicerae, Acetobacter thailandicus, Komagataeibacter oboediens, and Kozakia baliensis were significantly positively correlated (|r| ≥ 0.7, P < 0.05) with the majority of the compounds enriched on day 9. The genome-scale metabolic models (GSMMs) constructed for these core strains based on their metagenome-assembled genomes (MAGs) indicated that K. oboediens possessed complete carbon-nitrogen skeleton metabolism and flavor precursor synthesis capacity, while A. jinshanensis had urea cycle and cyanogenic glycoside degradation capacity, collectively forming the molecular basis for compounds accumulation. This research provides a theoretical foundation for the targeted design of synthetic microbial communities and the optimization of precision fermentation processes.

Key words: Shanxi aged vinegar; untargeted metabolomics; functional microorganisms; genome-scale metabolic models