Urinary mRNA Signatures as Predictors of Renal Function Decline in Patients With Biopsy-Proven Diabetic Kidney Disease.

The clinical manifestations of diabetic kidney disease (DKD) are more heterogeneous than those previously reported, and these observations mandate the need for the recruitment of patients with biopsy-proven DKD in biomarker research. In this study, using the public gene expression omnibus (GEO) repository, we aimed to identify urinary mRNA biomarkers that can predict histological severity and disease progression in patients with DKD in whom the diagnosis and histologic grade has been confirmed by kidney biopsy. We identified 30 DKD-specific mRNA candidates based on the analysis of the GEO datasets. Among these, there were significant alterations in the urinary levels of 17 mRNAs in patients with DKD, compared with healthy controls. Four urinary mRNAs-LYZ, C3, FKBP5, and G6PC-reflected tubulointerstitial inflammation and fibrosis in kidney biopsy and could predict rapid progression to end-stage kidney disease independently of the baseline eGFR (tertile 1 vs. tertile 3; adjusted hazard ratio of 9.68 and 95% confidence interval of 2.85-32.87, p < 0.001). In conclusion, we demonstrated that urinary mRNA signatures have a potential to indicate the pathologic status and predict adverse renal outcomes in patients with DKD.

A Comparison of Two-Stage Approaches Based on Penalized Regression for Estimating Gene Networks.

Graphical models are commonly used for illustrating gene networks. However, estimating directed networks are generally challenging because of the limited sample size compared with the dimensionality of an experiment. Many previous studies have provided insight into the problem, and recently, two-stage approaches have shown significant improvements for estimating directed acyclic graphs. These two-stage approaches find neighborhoods in the first stage and determine the directions of the edges in the second stage. However, although numerous methods to find neighborhoods and determine directions exist, the most appropriate method to use with two-stage approaches has not been evaluated. Therefore, we compared such methods through extensive simulations to select effective methods for the first and second stages. Results show that adaptive lasso is the most effective for both stages in most cases. In addition, we compared methods to handle asymmetric entries to estimate an undirected network. Some previous studies indicate that the method used to handle asymmetric entries does not affect performance significantly; however, we found that the selection of the handling method for such edges is a significant factor for finding neighborhoods when using adaptive lasso.