Comparative Metagenomics of Human Gut Microbiota Across Diets

Thesis Abstract

Dietary patterns exert a profound influence on the composition and functional potential of the human gut microbiota, yet comparative metagenomic analyses across diverse diets remain underexplored in large, well-controlled cohorts. This study addresses the problem of inconsistent associations between diet types and microbial community structure by integrating cross-sectional sampling with rigorous metagenomic profiling to elucidate taxonomic and functional shifts attributable to diet. The aim is to characterize how omnivorous, plant-based, and high-fat Western diets shape gut microbiota composition, gene repertoire, and predicted metabolic pathways, and to identify microbial signatures predictive of diet category. Specific objectives include (1) comparing alpha and beta diversity metrics across dietary groups; (2) identifying core and differential taxa using shotgun metagenomics; (3) quantifying functional potential and pathway abundances via eggNOG, KEGG, and MetaCyc annotations; (4) assessing associations between dietary intake quality indices and microbial functions using multivariate regression; and (5) exploring potential mediation by short-chain fatty acid biosynthesis pathways on host metabolic markers. A cross-sectional design will be employed in a multinational urban cohort of 600 adults aged 20–60 years, categorized into omnivores (n=200), plant-based vegetarians/vegans (n=200), and high-fat Western-style diet adherents (n=200). Fecal samples will be collected under standardized protocols and subjected to shotgun metagenomic sequencing on the Illumina NovaSeq platform, generating a minimum of 12 gigabases per sample to ensure comprehensive gene-space coverage. Dietary data will be captured using three 24-hour recalls and a validated food frequency questionnaire, enabling the computation of diet quality indices such as the Healthy Eating Index and macronutrient ratios. Additional host information, including age, sex, BMI, physical activity, medication use, and recent antibiotic exposure, will be recorded to adjust for confounding variables. Laboratory analysis will include fecal metabolomics for short-chain fatty acids in a subset (n=180) to triangulate functional inferences. Data analysis will proceed in three stages using established bioinformatics pipelines and statistical methods. First, sequencing reads will be quality-filtered, host-filtered, and assembled, with taxonomic profiling performed via Kraken2 and Bracken, and functional annotation assigned through HUMAnN3 against the UniRef90 gene catalog. Second, alpha diversity (Shannon, Simpson) and beta diversity (Bray-Curtis, UniFrac) metrics will be compared across dietary groups using ANOVA and PERMANOVA, adjusting for covariates. Differential abundance analyses will employ DESeq2 for taxa and advanced regression models for pathways, controlling false discovery rate at 5%. Third, multivariate modeling will integrate dietary variables, microbial features, and host metadata using partial least squares discriminant analysis (PLS-DA) and hierarchical Bayesian regression to quantify diet-driven effects and uncertainty. Mediation analyses will examine whether alterations in butyrate-producing pathways mediate associations between diet and host metabolic markers. Expected findings include higher relative abundance of Prevotella and fiber-degrading taxa in plant-based diets, enrichment of Bacteroides and bile-tolerant taxa in omnivores and Western diets, and differential enrichment of carbohydrate-active enzymes (CAZymes) corresponding to fiber type. Functional profiling is anticipated to reveal greater potential for short-chain fatty acid biosynthesis, particularly butyrate, in plant-based groups, contrasted with lipid metabolism pathways enriched in high-fat Western diets. These patterns are expected to correlate with improved metabolic biomarkers in plant-based groups, mediated by microbial-derived metabolites. The study will contribute to knowledge by delineating robust diet–microbiome–function relationships across diverse populations, establishing microbial signatures for dietary patterns, and informing precision nutrition strategies. The main conclusion posits that diet exerts a measurable and diet-specific impact on both microbial community composition and functional potential, with implications for host metabolism and health. Recommendations include the development of diet-tailored microbiome interventions to optimize SCFA production and metabolic outcomes, incorporation of metagenomic dashboards in dietary assessment practices, and longitudinal follow-up studies to investigate causal pathways and temporal stability of diet–microbiome associations.