Machine learning-based analysis identifies glucose metabolism-related genes ADPGK as potential diagnostic biomarkers for clear cell renal cell carcinoma.

Introduction: Clear cell renal cell carcinoma, with its high morbidity and mortality, is one of the more difficult diseases in the world and still lacks an effective therapeutic target. The primary way they break down glucose is through aerobic glycolysis, which leads to energy acquisition and synthesis of the material base required for cell growth. Although targeting glucose metabolism has driven the development of a variety of tumour therapies, the specific regulatory mechanisms remain unclear. Therefore, based on machine learning analysis algorithms, we analysed the correlation between glycometabolic pathways and ccRCC in the REACTOME database and verified the impact of the key gene Methods: We utilised a total of 89 gene collections of glucose metabolism pathways from the REACTOME (https://reactome.org/) database as the data base for our study. To uncover potential therapeutic target genes, we adopt three machine learning algorithms (LASSO, RF, and Boruta). We reassigned the 7 screened genes based on gene expression differences between cancer and paracancerous tissues, and applied an unsupervised consensus clustering algorithm to establish a typology based on the expression of glucose metabolism-related genes ( Results: We identified Discussion: Our results suggest that targeting the glucose metabolism pathway can kill ccRCC cells.