Food spoilage poses a significant challenge in the food industry, leading to economic losses and food insecurity. To address this issue, our study employs a comprehensive metagenomic and metatranscriptomic approach to characterize spoilage-related microbial communities in various food samples. By sequencing and analyzing both the DNA and RNA profiles, we capture the taxonomic composition and active metabolic functions associated with spoilage. Metagenome-assembled genomes (MAGs) are reconstructed to identify both culturable and non-culturable spoilage organisms, including novel and unannotated species. Taxonomic profiling reveals key spoilage taxa and tracks their progression under different storage conditions. Functionally, we map metabolic pathways involved in spoilage, identifying critical pathways responsible for the production of spoilage markers like biogenic amines and volatile compounds. Statistical analyses of microbial functions and taxonomic drivers highlight significant differences associated with storage conditions and spoilage stages. We integrate microbiome and sensory data, applying machine learning models to identify robust microbial and metabolic biomarkers for early spoilage detection. Building on these findings, we design synthetic microbial communities with antagonistic properties to inhibit spoilage processes.

In silico simulations validate the potential of these communities to outcompete spoilage organisms, providing a foundation for future experimental validation. This study demonstrates the potential of metagenomic approaches and synthetic biology to develop innovative, biologically-based strategies for controlling food spoilage and extending shelf life.