One of the most documented features about microbiomes is its great taxonomic variability, which is described in animals, like fish, as well as in humans. Despite this enormous inter and intra-species heterogeneity, at larger scale, microbial populations seem to exhibit a redundant functionality, in which phylogenetically related or unrelated taxa can share similar genes that accomplish for analogous general functions. At lower scale, biotic and abiotic factors such as those occurring in aquaculture production systems can actually disturb this balance, defining advantages and/or disadvantages in terms of performance. However, the bacterial specific functional contributions are difficult to be defined and measured.
Thus, the present investigation aimed to reinterpret the apparent functional redundancy of the microbiota from an applicative point of view using a Bayesian network (BN) approach. The bioinformatic analyses were performed using three experiments, which investigated the effects of different dietary fishmeal substitutions on anterior and posterior intestinal microbiota of gilthead sea bream. The construction of the BN was conducting using the updated version (V2) of SAMBA platform (Soriano et al., 2023; https://doi.org/10.3390/genes14081650), implemented with community detection method (Leiden method) with which the microbial population can be re-organized for the identification of strictly connected bacterial clusters. The specific functional profile of each cluster is then performed using KEGG pathways annotation and directly included in the BN. Our results allow the identification of key bacterial genera in the clusters (Fig. 1a), which trough the dietary modulation can lead to metabolic changes, such as those related to metabolites biosynthesis and lipid metabolism (Fig. 1b), that can also be part of the interaction with the host. These findings represent a specific and targeted approach of SAMBA. However, the platform can also be used to obtain a global and general organization of the microbial network associated with a compartment (e.g., anterior intestine, posterior intestine). Hence, the second aim of this study was to perform a meta-analysis, building a BN with SAMBA, merging different experiments, to identify the most representative connections between bacteria. Our results shown how the constant and variable fractions of microbial populations, including the CORE microbiota, are distributed in the network and interact with each other. All these outcomes represent the starting point to define a new reference of microbial organization, which can serve also as material for comparison between species or distinct body compartments.
This work was supported by MCIN with funding from European Union NextGenerationEU (PRTR-C17.I1) and by Generalitat Valenciana (THINKINAZUL/2021/024).