Integrated bioinformatics and machine learning reveal pan-apoptosis and immune infiltration signatures in diabetic nephropathy

Yu Liu¹Wenqian Lu²Zhicong Xiang³Yuli Shen⁴Baoyi Ni⁵Hequn Zou^2*Xiaofei Shao^6*

• ¹Luohu District Hospital of Traditional Chinese Medicine, Shenzhen, China
• ²The Chinese University of Hong Kong-Shenzhen, Shenzhen, China
• ³Shenzhen Luohu Hospital Group Luohu People's Hospital, Shenzhen, China
• ⁴The Second Affiliated Hospital, School of Medicine, The Chinese University of Hong Kong, Shenzhen, Shenzhen, China
• ⁵Heilongjiang University of Chinese Medicine, Harbin, China
• ⁶The Third Affiliated Hospital, Southern Medical University, Guangzhou, China

Objective: This study aimed to explore panoptosis-related genes and potential therapeutic drugs in DN. Methods: We downloaded DN datasets from the GEO database and identified differentially expressed genes (DEGs) through integrated differential expression analysis and weighted gene co-expression network analysis (WGCNA). The intersection between DN-related DEGs and panoptosis-related genes was obtained, and LASSO and SVM machine learning algorithms were applied to screen candidate biomarkers. The area under the receiver operating characteristic curve (AUC) was calculated for evaluation. Validation was performed using the merged dataset of GSE30529 and GSE4713. The CIBERSORT algorithm was used to assess immune cell infiltration, and Spearman correlation analysis was conducted to examine the association of biomarker genes. The Kidney Integrative Transcriptomics database was employed to explore the distribution of core genes across 12 cell populations. Potential drug molecules interacting with core genes were screened using the DSigDB database on the Enrichr platform, and molecular docking was performed using AutoDock Vina to evaluate binding affinity. The qRT-PCR was used to validate the expression of these hub mitochondria-related genes. Results: Analysis of the DN dataset yielded 17 intersecting genes. Gene Ontology (GO) and Kyoto Encyclopedia of Genes and Genomes (KEGG) enrichment analyses revealed that these genes were significantly associated with immune and inflammatory responses, pyroptosis, extrinsic apoptosis, necroptosis, and related pathways. Using LASSO and SVM machine learning algorithms, eight candidate biomarkers were identified: CD44, CRIP1, CEBPB, TNFRSF1B, CAV1, IGF1, GZMB, and LY96. ROC curve analysis demonstrated that these biomarkers had strong diagnostic value for DN patients. Further investigation into immune infiltration in DN samples using CIBERSORT showed that core genes were closely related to dendritic cells (resting), macrophages (M1), mast cells (activated), neutrophils, T cells (CD4 memory activated, CD4 memory resting, CD8, and gamma delta). Drug screening via DSigDB on Enrichr identified imatinib as a significantly enriched drug interacting with core genes, and molecular docking confirmed its strong binding affinity. Conclusion: Through comprehensive bioinformatics approaches, this study identified CD44, CRIP1, CEBPB, TNFRSF1B, CAV1, IGF1, GZMB and LY96 as potential diagnostic biomarkers for DN, providing new insights into disease diagnosis.