Document resource
Background Non-alcoholic fatty liver disease (NAFLD) is the most common chronic liver disease worldwide, yet effective treatments remain limited. Gut microbiota regulates metabolic homeostasis via the gut-liver axis, but inter-individual heterogeneity obscures causal bacterial strains. This study integrates machine learning analysis of public microbiome data with droplet microfluidic technology to systematically identify characteristic gut microbiota and design an artificial consortium of microbiota for NAFLD intervention.Methods NAFLD gut microbiome cohorts were retrieved from public databases. Multiple classifiers and SHAP analysis identified core microbial features. Target strains were isolated from healthy human feces using droplet microfluidic technology and assembled into an artificial consortium of microbiota (ACM), whose stability was validated through in vitro co-culture. The artificial consortium was then co-cultured with NAFLD fecal microbiota, followed by 16S rRNA sequencing and untargeted metabolomics. Therapeutic efficacy was evaluated in MCD diet-induced NAFLD mice. The overall workflow is illustrated in Figure 1 ( IDDF2026-ABS-0137 Figure 1. Integrated Workflow for Rational Design and Validation of Artificial Consortium of Microbiota (ACM) Targeting NAFLD).Results Machine learning analysis of public NAFLD cohorts, combined with SHAP, identified characteristic gut microbiota ( IDDF2026-ABS-0137 Figure 2. Machine learning-driven prioritization of target gut microbiota for NAFLD intervention). Target strains were isolated from healthy human feces using droplet microfluidics and assembled into ACM, whose stability was validated through co-culture. ACM was co-cultured with NAFLD fecal microbiota. 16S rRNA sequencing revealed that ACM treatment altered microbial composition, preserved alpha diversity, and restored beneficial taxa (IDDF2026-ABS-0137 Figure 3. 16S rRNA sequencing and diversity analysis reveal gut microbiota restructuring by Artificial Consortium of Microbiota (ACM) in NAFLD fecal culture system). Untargeted metabolomics revealed ACM-driven metabolic remodeling, with key pathways including pantothenate and CoA biosynthesis, glycine, serine and threonine metabolism, and arginine and proline metabolism (IDDF2026-ABS-0137 Figure 4. Untargeted metabolomics coupled with pathway enrichment analysis identifies ACM-driven metabolic remodeling in NAFLD fecal culture system). In MCD diet-induced NAFLD mice, biochemical and histological assessment demonstrated that ACM reduced liver index, serum ALT/AST, and hepatic lipid accumulation, confirming its therapeutic efficacy (IDDF2026-ABS-0137 Figure 5. Biochemical analysis and morphological assessment demonstrate therapeutic efficacy of Artificial Consortium of Microbiota (ACM) in MCD diet-induced NAFLD mice).Conclusions This study developed a machine learning-guided and droplet microfluidic-based isolation pipeline for the construction of artificial consortia of microbiota, which ameliorates NAFLD progression by reshaping gut microbial communities and activating protective metabolic pathways. This suggests that precisely engineered microbiota represent a novel therapeutic approach for NAFLD patients.Abstract IDDF2026-ABS-0137 Figure 1Abstract IDDF2026-ABS-0137 Figure 2Abstract IDDF2026-ABS-0137 Figure 3Abstract IDDF2026-ABS-0137 Figure 4Abstract IDDF2026-ABS-0137 Figure 5