Multi-omics analysis reveals the aging mechanism during Baobaoqu storage for the production of Wuliangye Baijiu.

The quality of Baobaoqu directly influences the quality of Baijiu. In this study, metatranscriptomics, metaproteomics and metabolomics were utilized, combining machine learning with a multi-layer perception neural network model and the Shapley additive explanations approach, to analyze the metabolic mechanisms in Baobaoqu of varying quality. The results indicated that normal Baobaoqu contained more flavor compounds, while premium Baobaoqu contained more active microorganisms and enzymes. During storage, the number of differential proteins between premium and normal Baobaoqu increased, reaching 504 and 77, respectively. In premium Baobaoqu, cellulase expression gradually decreased, while that of ethanol metabolic enzymes increased; the opposite trend was observed in normal Baobaoqu. These differences were primarily attributed to variations in the proportions of Aspergillus and Thermoascus. The findings of this study indicate that a suitable combination of premium and normal Baobaoqu is necessary and provides a scientific basis for optimizing the Baobaoqu aging conditions and enhancing Baijiu quality.