Microbiome subtyping and network analysis in inflammatory bowel disease for clinical and multi-omics characterization

Abstract

Inflammatory bowel disease （IBD）, including Crohn’s disease （CD） and ulcerative colitis （UC）, is associated with gut microbial dysbiosis, altered microbial metabolism and aberrant host immune responses. However, due to the heterogeneity of IBD, patients with same diagnosis can exhibit different microbial compositions, inflammatory responses and clinical severity. In this study, taxonomy-based subtyping was conducted using fecal shotgun metagenomic profiles of a Korean IBD cohort to address this heterogeneity, identifying distinct subtypes within CD and UC. Specifically, subtypes with lower diversity showed distinct microbial compositions compared to non-IBD controls （NM）, in particular, depletion of butyrate-producing taxa and enrichment of lactate-associated bacteria. Furthermore, network analysis and community detection demonstrated that these depleted and enriched bacteria were clustered into distinct co-abundant communities, which were shared across CD and UC, suggesting core community structures in IBD microbiome. To connect microbial communities with host phenotypes in IBD, associations of these communities with subtype information, fecal metabolome, host colonic transcriptome and clinical metadata were subsequently examined. Notably, these communities were associated with fecal metabolites, including bile acids, lactate and butyrate, as well as host inflammatory pathway signatures and clinical indicators of disease activity. Together, these findings suggest that taxonomy-based subtypes and co-abundant communities using microbiome can provide an interpretable framework for stratifying IBD heterogeneity and linking taxonomic dysbiosis to mechanistic and clinical interpretation.