Year: 2026
Type: preprint
Abstract: Background Public shotgun metagenome repositories offer a valuable resource for identifying gut microbial functional targets, but their reuse is challenging because of heterogeneous cohort structures, incomplete metadata, high-dimensional taxonomic features, and incomplete functional annotation. This study aimed to develop a computational framework to prioritize reproducible microbial functional modules across multiple public metagenomic cohorts. Results Processed stool metagenome data from 11 public cohorts comprising 2,255 samples were analyzed. Functional modules provided a more compact and less heterogeneous feature space than taxonomic features. Across cohorts, three short-chain fatty acid (SCFA)-related modules—CENTFERM-PWY, PWY-6590, and P461-PWY—showed consistent multi-cohort evidence and met the predefined meta-FDR and direction-consistency criteria. The Functional Target Priority Score ranked CENTFERM-PWY and PWY-6590 as the top candidates, followed by tryptophan-related modules, while GABA-related pathways remained exploratory. In an independent held-out cohort, 28 of 31 tested modules showed concordant directions, corresponding to a 90.3% direction agreement fraction. Conclusions This multi-cohort framework identifies SCFA-related microbial fermentation modules as the most reproducible computational candidates for gut microbial functional targeting. The findings provide a data-driven prioritization strategy for public metagenomic resources while highlighting the need for experimental and clinical validation before biological or interventional conclusions can be made.
Source: Research Square
Language: English
Topic: Gut microbiota and health
Subfield: Molecular Biology
Domain: Life Sciences
Open Access status: green
OpenAlex ID: W7236613673
