Back

Seven gut microbiome types across two Japanese cohorts defined by co-abundance-based microbial guilds and associated with host physiology

NAKAYAMA, J.; Yoden, M.; Uchikawa, A.; Tanaka, M.; Tamura, S.; Shinoda, A.; Momoda, R.; Kameyama, K.; Inoue, T.; Watanabe, H.; Tanaka, Y.

2026-09-08 microbiology
10.64898/2026.09.07.749831 bioRxiv
Show abstract

Intensive efforts have been made to capture structural features of the gut microbiome for assessing microbiota status in relation to host health. However, gut microbiome function emerges not only from individual microbial taxa but also from their organization into ecological guilds, complicating microbiome structure-function relationships. To address this challenge, we applied a co-abundance-based analytical framework to identify gut microbiome types (GMTs) across a discovery cohort (D-cohort) and an independent validation cohort (V-cohort) of Japanese adults. Using paired fecal samples collected one month apart from 116 participants in the D-cohort, we identified 27 co-abundance groups (CAGs) based on 16S rRNA amplicon sequence variants. These CAGs defined seven GMTs (MT-1 to MT-7) with characteristic short-chain fatty acid and bile acid profiles. The GMTs showed temporal stability, with more than 70% of participants remaining in the same GMT across sampling points. The 27-CAG framework classified 404 participants in the independent V-cohort into the seven GMTs, which represented reproducible microbial ecological states across cohorts. MT-4 exhibited dysbiosis-like features, including reduced microbial diversity, depletion of Faecalibacterium and butyrate, and enrichment of inflammation-associated taxa, and was associated with higher alcohol consumption and elevated {gamma}-glutamyl transferase levels. In contrast, Bifidobacterium-enriched MT-5 was associated with favorable renal function profiles. Collectively, co-abundance-defined microbial guilds capture coherent ecological and functional states of the gut microbiota, providing a transferable framework for linking gut microbiome organization to host physiology beyond conventional taxonomy-based classification.

Matching journals

The top 5 journals account for 50% of the predicted probability mass.

50% of probability mass above

"Similar papers" are the closest papers from that journal in the model's embedding space. They show what the match is built on, but the ranking comes mostly from a classifier over the whole training set, not from these examples alone.