Machine-learning assisted discovery unveils novel interplay between gut microbiota and host metabolic disturbance in diabetic kidney disease

Clinical Snapshot

Key Findings

• P Value: Significance levels not specified in abstract

• Effect Size: Machine learning accuracy 0.72-0.78 across different disease comparisons

• Primary Outcome: Identification of 13 gut microbiota biomarkers discriminating between disease groups

• Nnt Or Sensitivity: Area under curve 0.79-0.86 for diagnostic discrimination

• Confidence Interval: Not reported in abstract

Clinical Application

Requires specialised microbiota sequencing facilities and machine learning expertise, limiting immediate clinical application Relevant to Australian diabetes management given high T2DM prevalence; would require validation in Australian population and integration with existing RACGP diabetes guidelines Adults with type 2 diabetes, diabetic kidney disease, or chronic kidney disease

Abstract

Diabetic kidney disease (DKD) is a serious healthcare dilemma. Nonetheless, the interplay between the functional capacity of gut microbiota and their host remains elusive for DKD. This study aims to elucidate the functional capability of gut microbiota to affect kidney function of DKD patients. A total of 990 subjects were enrolled consisting of a control group (n = 455), a type 2 diabetes mellitus group (DM, n = 204), a DKD group (n = 182) and a chronic kidney disease group (CKD, n = 149). Full-length sequencing of 16S rRNA genes from stool DNA was conducted. Three findings are pinpointed. Firstly, new types of microbiota biomarkers have been created using a machine-learning (ML) method, namely relative abundance of a microbe, presence or absence of a microbe, and the hierarchy ratio between two different taxonomies. Four different panels of features were selected to be analyzed: (i) DM vs. Control, (ii) DKD vs. DM, (iii) DKD vs. CKD, and (iv) CKD vs. Control. These had accuracy rates between 0.72 and 0.78 and areas under curve between 0.79 and 0.86. Secondly, 13 gut microbiota biomarkers, which are strongly correlated with anthropometric, metabolic and/or renal indexes, concomitantly identified by the ML algorithm and the differential abundance method were highly discriminatory. Finally, the predicted functional capability of a DKD-specific biomarker, Gemmiger spp. is enriched in carbohydrate metabolism and branched-chain amino acid (BCAA) biosynthesis. Coincidentally, the circulating levels of various BCAAs (L-valine, L-leucine and L-isoleucine) and their precursor, L-glutamate, are significantly increased in DM and DKD patients, which suggests that, when hyperglycemia is present, there has been alterations in various interconnected pathways associated with glycolysis, pyruvate fermentation and BCAA biosynthesis. Our findings demonstrate that there is a link involving the gut-kidney axis in DKD patients. Furthermore, our findings highlight specific gut bacteria that can acts as useful biomarkers; these could have mechanistic and diagnostic implications.

References

1. 1.Wu, I-W., Liao, Y-C., Tsai, T-H., Lin, C-H., Shen, Z-Q., Chan, Y-H., Tu, C-W., Chou, Y-J., Lo, C-J., Yeh, C-H., Chen, C-Y., Pan, H-C., Hsu, H-J., Lee, C-C., Cheng, M-L., Sheu, W. H-H., Lai, C-C., Sytwu, H-K., & Tsai, T-F. (2025). Machine-learning assisted discovery unveils novel interplay between gut microbiota and host metabolic disturbance in diabetic kidney disease. Gut Microbes. https://doi.org/10.1038/s41467-024-52097-9