PhD Defense:

Comprehensive study of the microbiome of fermented vegetables using integrative approaches

The development of high-throughput sequencing has sparked renewed interest in studying various microbiomes. Among them, the microbiome of fermented vegetables is particularly intriguing due to its spontaneous establishment and high diversity. This poses challenges in control of production and organoleptic and nutritional properties. The management of food fermentation still relies heavily on empirical knowledge, as understanding the dynamics of microbial communities and the metabolic networks producing safe and nutritious products still needs to be explored. The open science paradigm, especially the availability of numerous datasets, provides an opportunity to characterize this microbiome more in-depth. This thesis presents various integrative methods for analyzing the microbiome of fermented vegetables, incorporating diverse public data to characterize and compare microbiomes across fermented vegetables comprehensively. Firstly, we investigated the succession dynamics of bacterial communities during the fermentation of vegetables. Therefore, we designed a network-based approach to compare ten public metataxonomic 16S datasets, including 931 samples, and targeting different fermented vegetables throughout time. Networks for individual datasets are integrated into a core network, highlighting significant associations. This method sheds light on the dynamics of vegetable fermentation by characterizing the processes of community succession among different bacterial assemblages. On a secondary level, we used a multi-marker metataxonomic approach to study the influence of processing factors on the bacterial communities during the fermentation.We then investigated the relationship between the core microbiota and the core metabolic pathways of fermented vegetables. To this end, we comprehensively analyzed the microbial diversity, taxonomic composition, and metabolic profiles of nine independent datasets, including 141 samples. Using a reproducible analytical workflow with a read-based approach, we identified a core set of microbial species and strains. We linked them to a set of shared metabolic functions that may represent a network of activities important during the fermentation of vegetables.

Jury members

• Thibault Nidelet, Chargé de recherche, Centre INRAE Occitanie-Montpellier
• Emmanuelle Becker, Directrice de recherche, Irisa/Inria Rennes-Bretagne Atlantique
• Jérôme Mounier, Professeur des universités, Université de Bretagne Occidentale
• Claudine Médigue, Directrice de recherche, Genoscope

Supervised by

Hélène Chiapello et de Stéphane Chaillou, et avec l’encadrement de Florence Valence et Michel-Yves Mistou