Every macroorganisms are associated with microorganisms forming the host-microbiota. Host-organisms are highly dependent on these symbioses for their survival and fitness. Theories about holobionts suggest that the microbiome is faster to adapt than the host, further emphasizing the need to focus on this cornerstone of ecosystems studies. A major concern is to improve the knowledge to explain the complexity of the microbiota communities as well as the mechanisms underlying the general ecosystem functionalities. To better understand the Arabidopsis thaliana microbiota, we developed, in the framework of the systems-based theory, a network modelling approach. Multi-omics data were used to improve our understanding of the microorganism coexistence within the microbiota. Functional annotation, metabolic network modeling and biological pathway completion inform on potential interaction mechanisms at the metabolic scale. The hypothesis according to which microbial communities gather preferentially in order to optimize functional complementarity can explain the metabolic richness and diversity of the microbiome observed in the plant holo-biont.