Integrated multi-omics analysis unveils microbiota-metabolite-host interactions and novel biomarkers for early diabetic kidney disease diagnosis - PubMed
5 hours ago
- #multi-omics
- #diabetic kidney disease
- #machine learning
- Integrated multi-omics analysis reveals microbiota-metabolite-host interactions and novel biomarkers for early diabetic kidney disease (DKD) diagnosis.
- Focuses on East Asian population due to distinct genetic, environmental, and lifestyle factors influencing DKD.
- Uses Mendelian randomization (MR) analysis to explore causal relationships between microbiota, metabolites, and DKD.
- Identifies significant associations between specific microbiota taxa (e.g., Haemophilus-A, TM7x) and metabolites (e.g., tyrosine, glutamine) related to DKD.
- Clinical samples show microbial dysbiosis in DKD patients, including increased Klebsiella and reduced Faecalibaculum and Dubosiella.
- Metabolomic profiling reveals alterations in branched-chain amino acids (BCAAs) and fatty acids in DKD.
- Machine learning models achieve over 90% accuracy in distinguishing type 2 diabetes mellitus (T2DM) from DKD.
- Suggests multi-omics integration and ML could improve early DKD detection and personalized management in East Asian populations.