Integrating Bioinformatics and Machine Learning to Identify Glucose Metabolism-Related Biomarkers with Diagnostic and Prognostic Value for Patients with Kidney Renal Clear Cell Carcinoma.

Background: Glucose metabolism plays a critical role in the development and progression of kidney renal clear cell carcinoma (KIRC). This study aimed to identify glucose metabolism-related biomarkers (GRBs) and therapeutic targets for KIRC diagnosis and prognosis using bioinformatics and machine learning.

Methods: Gene expression data from the Gene Expression Omnibus (GEO) and The Cancer Genome Atlas (TCGA) databases, along with glucose metabolism-related genes from multiple sources, were analyzed. Differentially co-expressed glucose metabolism-related genes (DCGLGs) were identified through differential expression analysis and weighted gene co-expression network analysis. Functional enrichment analysis and protein-protein interaction network construction were performed on the DCGLGs. Machine learning algorithms identified GRBs, evaluated for diagnostic value via receiver operating characteristic (ROC) curve analysis. Further analyses included enrichment, immune infiltration, drug sensitivity, clustering, and Kaplan-Meier survival analysis of GRBs.

Results: Among 884 glucose metabolism-related genes, 39 DCGLGs were identified. Ten GRBs were highlighted, all exhibiting high diagnostic value (area under the ROC curve (AUC) >0.85). GRBs were linked to immune cell infiltration, including endothelial cells and CD4+ T cells. Drug sensitivity analysis revealed significant correlations between Phosphofructokinase platelet (PFKP) and multiple chemotherapeutic agents. Clustering based on GRBs stratified patients into two clusters, with cluster 2 showing poorer prognosis. Kaplan-Meier survival analysis validated the prognostic significance of GRBs.

Conclusions: GRBs, including PFKP, pyruvate dehydrogenase kinase 1 (PDK1), and solute carrier family 2 member 1 (SLC2A1), demonstrated strong diagnostic and prognostic potential. PFKP emerged as a key therapeutic target, offering novel insights into predictive and treatment strategies for KIRC.