Diabetic kidney disease (DKD) is the most common complication of diabetes and a leading cause of chronic kidney disease that frequently leads to end-stage renal disease (ESRD). The pathogenesis of DKD is complex and is not fully understood. This study was designed to identify key targets for DKD diagnosis and explore the underlying molecular mechanisms.DKD-specific clusters were selected from single-cell datasets. Gene modules were identified using hairpin-dynamic weighted gene co-expression network analysis (hdWGCNA). Multiple machine learning algorithms were applied to model and screen hub genes from two bulk datasets. Rat model of DKD was built using optical microscopes to observe the histopathological changes in the kidney by HE, PAS, and Masson staining. The expression of RASGRP3, PDE3B, and CD247 in DKD-Rat was verified by RT-PCR, and the expression of RASGRP3, PDE3B, and CD247 in the serum samples of DKD patients was verified by ELISA. The results of sex and age, RASGRP3, PDE3B, CD247 were calculated by multivariate logistic regression analysis.Three hub genes were obtained through screening single-cell and two bulk datasets. In-depth exploration of the potential molecular mechanisms of the hub genes was conducted using gene set variation analysis (GSVA), immune infiltration analysis, and single-cell correlation analysis. Receiver operating characteristic (ROC) curve confirmed a high diagnostic value of the hub biomarkers, and a high-efficiency diagnostic model was constructed and mutually validated in the two datasets. We found that damaged tubular number and interstitial fibrotic percentage were significantly increased in DKD rat. As shown by HE, PAS and Masson staining, the mRNA levels of PDE3B and CD247 were markedly upregulated in DKD rat compared with those in the control group. Lower expression levels of RASGRP3 mRNA were manifested in DKD. The levels of RASGRP3, PDE3B, CD247 in DKD patients by ELISA were statistically significant (p < 0.05). PDE3B and CD247 had an AUC value greater than 0.9,RASGRP3 had an AUC value greater than 0.7.This study identified 3 T cell-related hub biomarkers, providing references for the early diagnosis of DKD and changes in T cells during DKD progression.