The human gut microbiota in IBD, characterizing hubs, the core microbiota and terminal nodes: a network-based approach

BACKGROUND: Dysbiosis, an imbalance in the bacterial composition of the human gut microbiota, is linked to inflammatory bowel disease (IBD). Advances in biological techniques have generated vast microbiota datasets, presenting both opportunities and challenges for clinical research in that field. Network theory offers powerful tools to analyze these complex datasets. METHODS: Utilizing genetically unrelated individuals from the Kiel IBD-KC cohort, we compared network properties of the gut microbiota between patients with inflammatory bowel disease (IBD, n = 522) and healthy controls (n = 365), and between Crohn's disease (CD, n = 230) and Ulcerative Colitis (UC, n = 280). Correlation-based microbial networks were constructed, with genera as nodes and significant pairwise correlations as edges. We used centrality measures to identify key microbial constituents, called hubs, and suggest a network-based definition for a core microbiota. Using Graphlet theoretical approaches, we analyzed network topology and individual node roles. RESULTS: Global network properties differed between cases and controls, with controls showing a potentially more robust network structure characterized by e.g., a greater number of components and a lower edge density. Local network properties varied across all groups. For cases and both UC and CD, Faecalibacterium and Veillonella, and for unaffected controls Bacteroides, Blautia, Clostridium XIVa, and Clostridium XVIII emerged as unique hubs in the respective networks. Graphlet analysis revealed significant differences in terminal node orbits among all groups. Four genera which act as hubs in one state, were found to be terminal nodes in the opposite disease state: Bacteroides, Clostridium XIVa, Faecalibacterium, and Subdoligranulum. Comparing our network-based core microbiota definition with a conventional one showed an overlap in approximately half of the core taxa, while core taxa identified through our new definition maintained high abundance. CONCLUSION: The network-based approach complements previous investigations of alteration of the human gut microbiota in IBD by offering a different perspective that extends beyond a focus solely on highly abundant taxa. Future studies should further investigate functional roles of hubs and terminal nodes as potential targets for interventions and preventions. Additionally, the advantages of the newly proposed network-based core microbiota definition, should be investigated more systematically.