Fermented foods and beverages are among the oldest human biotechnologies and produce microbial metabolites that can modulate the gut microbiota, immune responses, and host metabolism. Growing prevalence of non-communicable chronic diseases such as obesity, type 2 diabetes, and chronic inflammation has motivated searches for microbiome-based interventions. In this context, fermented beverages represent promising reservoirs of next-generation probiotic strains and functional microbial consortia. The authors set out to apply a systematic, genome-resolved metagenomic pipeline to traditional fermented beverages to (1) characterize biosynthetic diversity relevant to host health and (2) use genome-scale metabolic modeling to inform the rational, in silico design of a synthetic probiotic community.
The study focused on milk-based kefir, kombucha, and two traditional Mexican fermented beverages, pozol and pulque. These beverages have previously been studied with high-depth shotgun metagenomics that produced high-quality genomic resources. The authors applied a bioprospecting pipeline to metagenomic assemblies to recover metagenome-assembled genomes (MAGs) suitable for downstream functional mining and metabolic modelling. Specific computational resources, assembly parameters, and the total number of MAGs recovered were described in the original manuscript and are available through the study's data and code links.
The authors conducted targeted genomic searches across recovered MAGs for genes and pathways linked to functions of interest: biosynthesis of B-group vitamins, production of short-chain fatty acids (SCFAs), biosynthetic gene clusters for natural products with potential antimicrobial or signaling functions, and carbohydrate-active enzymes (CAZymes) that may enhance utilization of starch and dietary fiber after intestinal colonization.
This focused mining aimed to identify community members that could contribute metabolic functions relevant to host nutrition and gut ecology. CAZymes were prioritized for their capacity to degrade complex carbohydrates, which could improve fiber and starch accessibility in the gut. Detection of vitamin biosynthesis pathways and SCFA-producing genes helped nominate strains likely to produce compounds that modulate host metabolism and immune function.
From the set of high-quality MAGs the authors generated genome-scale metabolic models to predict each organism's metabolic capabilities and potential roles in a mixed community. These models were used to identify candidate microbes predicted as central producers of secondary metabolites implicated in pathogen control and other community-level functions. The modeling approach enabled evaluation of metabolic complementarity, potential cross-feeding interactions, and the capacity of individual taxa to supply key metabolites to partners.
Building on genome mining and metabolic modeling results, the authors performed a rational, in silico assembly of a six-member synthetic microbial community. The selected combination of MAG-derived models was predicted to display stable, cooperative growth and high overall metabolic functionality, based on modelled nutrient exchanges and growth compatibility. The synthetic community represents a theoretical construct designed to maximize complementary biosynthetic activities identified in the beverage microbiomes.
Cross-feeding analysis of the designed community revealed shared metabolites and elements exchanged between members. Iron emerged as one of the most widely shared elements among community members. In this analysis, Priestia flexa—a taxon recovered from pozol—was predicted to act as a major donor of compounds involved in iron transport and to serve as a stabilizing element within the synthetic community. These results highlight how single taxa within fermented-beverage microbiomes can provide keystone functions that support community stability and cooperative metabolism.
The study demonstrates a pipeline that couples genome-resolved metagenomics, targeted functional mining, and genome-scale metabolic modeling to move from natural microbial diversity to a rationally designed, theoretical probiotic consortium. By identifying metabolic producers and predicted cross-feeding networks, this approach provides a framework for prioritizing strains for experimental validation and for the systematic exploitation of fermented-beverage microbial diversity in biomedical applications. The authors position this work as opening avenues for next-generation probiotic and functional consortium development aimed at microbiome modulation and host health.
The preprint notes that data and code supporting the analyses are provided via linked repositories. Funding sources declared include a challenge-based research funding program and university research funds. The work is presented as a preprint and has not been peer reviewed; details such as exact MAG counts, assembly metrics, and modeling parameters are reported in the full manuscript and external links. The study establishes an in silico design and prediction platform; experimental validation of the designed community and functional effects in vivo or in vitro would be required to confirm predicted activities and stability.