We have traditionally studied microbial physiology and functions in reductionist ways –growing microbes in isolation under well-controlled conditions. It is thus not surprising that functions relevant for interactions with other microbes, predators (e.g. phages) or hosts have remained understudied, despite their relevance in nature and the strong selection pressure to evolve. It is tempting to speculate that such “interaction-related” functions constitute a significant part of the remaining “unknowns” in microbial genomes. For the human gut microbiota these functions are of particular interest, as microbes form complex stable communities in dynamic environments, where they interact with the host and other community members, and fend off incoming microbes from food. From work on model enterobacteria, it has become evident that phylogenetically similar species and strains of the same species compete for the same niche, and use different ways to eliminate each other. In preliminary work using a strain collection of one of the most abundant and prevalent bacterial species in the human gut, Phocaeicola vulgatus, and a quantitative high-throughput interaction assay, we observed many intraspecies interactions and identified new putative toxins. In this porject, we aim to investigate the mechanisms behind the biosynthesis, maturation, export, immunity and mode-of-action of a new toxin (bacteriocin) of P. vulgatus, using molecular and biochemical approaches. We also want to expand our discovery pipeline for identification of diffusible toxins by increasing the size of our strain collections and expanding to other abundant and prevalent Bacteroides (B. uniformis), and by utilizing our established ability to construct and deconvolute genome-wide mutant libraries in the Bacteroidota phylum. Thereby we envision to provide a paradigm that can be applied to other gut microbes, and contribute to dissecting interactions that shape the human gut microbiome, and are integral for its stability, individuality and colonization resistance.