Deep longitudinal phenotyping of a single individual enables host–microbiome metabolic modeling over 3.6 years
This study examines a single subject (LS) experiencing severe episodic colonic inflammation and associated gut microbiome dysbiosis. Metagenomic sequencing of LS's stool samples at 51 time points over 3.6 years provided a time series of gut microbial species' relative abundances, from which we generated time-point specific microbiome models. Flux balance analysis was used to predict metabolic phenotypes across the time series. The microbial abundances and predicted fluxes were compared with clinical inflammation biomarkers [serum C-reactive protein (CRP); stool lactoferrin, calprotectin, lysozyme, and secretory immunoglobulin A]. Each of these 51 microbiome models were integrated with a time-point specific male whole-body metabolic model personalized based on LS's weight, blood glucose, and cholesterol. A healthy average (HeAve) community model from 34 individuals provided a reference. Our models predicted methane overproduction from high Methanobrevibacter smithii levels, coinciding with peak CRP and calprotectin. We found that Escherichia coli plays a central role in shaping the predicted metabolome and may promote inflammation through trimethylamine N-oxide and formaldehyde production, out-competing Desulfovibrio piger and Akkermansia muciniphila via active methyl and sulfur cycles. At a later time point, LS experienced a major shift in gut Proteobacteria ecology: E. coli, over-abundant at 675 × HeAve, dropped nearly 80-fold in one month, while Morganella morganii rose by nearly eight times, which coincided with a spike in LS's lactoferrin and lysozyme, a rise in species including Klebsiella pneumoniae and Faecalibacterium prausnitzii, a change in predicted metabolic profile, and increased alpha diversity. We also noted a correlation between postulated inflammation periods and predicted lithocholic acid levels. This work extends two previous studies of 7 and 27 of these time points. Our analysis demonstrates the value of pairing multi-omics time-series data with systems biology approaches in an n-of-1 design, resolving how microbiome composition dynamics shape host–microbiome metabolism and inflammation.
Authors
- Rob D. Knight (ORCID: https://orcid.org/0000-0002-0975-9019)
- Bernhard Ørn Palsson (ORCID: https://orcid.org/0000-0003-2357-6785)
- Larry L. Smarr (ORCID: https://orcid.org/0000-0003-2003-7700)
- Ines Thiele (ORCID: https://orcid.org/0000-0002-8071-7110)
- Xin Fang (ORCID: https://orcid.org/0000-0002-2078-2607)
- Filippo Martinelli (ORCID: https://orcid.org/0000-0002-6635-6793)
- Anna Sheehy
Institutions
- Ollscoil na Gaillimhe – University of Galway (IE)
- Hong Kong University of Science and Technology (HK)
- University College Cork (IE)
- Novo Nordisk Foundation (DK)
- University of California San Diego (US)
- Bristol-Myers Squibb (United States) (US)
- APC Microbiome Institute (IE)
- Technical University of Denmark (DK)
Publication Details
- Journal
- Gut Microbes
- Published
- 2026-11-19
- DOI
- https://doi.org/10.1080/19490976.2026.2757368
- Primary Topic
- Gut microbiota and health
- Type
- article
- Field-Weighted Citation Impact
- 0.00