Back

Food Microbiome Metabolic Modules (F3M), a tool suite for functional profiling of food microbiomes

Tap, J.; Mohellibi, N.; Tinsley, C.; Loux, V.; Chaillou, S.

2025-07-31 microbiology
10.1101/2025.07.31.667869 bioRxiv
Show abstract

The study of microbial metabolic interactions within food microbiomes represents a key scientific approach for improving the quality and health benefits of food. In such studies, methods based on gene expression levels (metatranscriptomic) analysis are promising. However specific tools are required to overcome the challenges posed by food microbiomes, in particular the high variability of microbiomes between samples and the difficulty of automatically inferring the annotation of metabolic functions across taxa. To adress this gap, we present the Food Microbiome Metabolic Modules (F3M) tool suite, which comprises (1) a curated database containing about 1,985 functional genes representing key fermentative metabolic reactions in food microbiomes, (2) a F3M Builder for generating F3M-annotated gene catalogs and mapping of metatranscriptomic reads, and then (3) an F3M R package to parse and aggregate gene expression data by taxonomic and functional categories for downstream analysis. The F3M taxonomy is organized according to the Genome Taxonomy Database (GTDB) nomenclature, whereas the F3M functional repertoire is structured hierarchically into 183 metabolic modules, which enable multi-scale analysis of inter-organism metabolic interactions and meaningful fermentative outputs (e.g., primary alcohols, acetate). Notably, a dedicated redox module captures oxido-reduction mechanisms and NADH-dependent pathways central to fermentation, while an uptake module complements the metabolic pathways to trace potential metabolite exchanges across taxa. Together, the F3M suite provides a robust framework for uncovering functional dynamics within food microbiomes. The F3M tool suite is available as open-source.

Matching journals

The top 8 journals account for 50% of the predicted probability mass.

50% of probability mass above

"Similar papers" are the closest papers from that journal in the model's embedding space. They show what the match is built on, but the ranking comes mostly from a classifier over the whole training set, not from these examples alone.