Clear cell renal cell carcinoma (ccRCC) is a major histological subtype of renal cell carcinoma and can be clinically divided into four stages according to the TNM criteria. Identifying clinical stage‐related genes is beneficial for improving the early diagnosis and prognosis of ccRCC. By using bioinformatics analysis, we aim to identify clinical stage‐relevant genes that are significantly associated with the development of ccRCC. First, we analyzed the gene expression microarray data sets: GSE53757 and GSE73731. We divided these data into five groups by staging information—normal tissue and ccRCC stages I, II, III, and IV—and eventually identified 500 differentially expressed genes (DEGs). To obtain precise stage‐relevant genes, we subsequently applied weighted gene coexpression network analysis (WGCNA) to the GSE73731 dataset and KIRC data from The Cancer Genome Atlas (TCGA). Two modules from each dataset were identified to be related to the tumor TNM stage. Several genes with high inner connection inside the modules were considered hub genes. The intersection results between hub genes of key modules and 500 DEGs revealed UBE2C, BUB1B, RRM2, and TPX2 as highly associated with the stage of ccRCC. In addition, the candidate genes were validated at both the RNA expression level and the protein level. Survival analysis also showed that 4 genes were significantly correlated with overall survival. In conclusion, our study affords a deeper understanding of the molecular mechanisms associated with the development of ccRCC and provides potential biomarkers for early diagnosis and individualized treatment for patients at different stages of ccRCC.
Background: Early detection of colorectal cancer (CRC) is crucial for improving patient outcomes. Cell-free DNA (cfDNA) analysis has emerged as a promising non-invasive approach for cancer detection. This study aims to develop a machine learning algorithm leveraging cfDNA fragmentomic features to accurately detect CRC. Methods: 573 individuals from Sir Run Run Shaw Hospital, two community healthcare centers and three additional medical centers, were collected between April 1, 2023, and December 12, 2025. Participants were divided into training, internal validation, and external validation cohorts. A variety of cfDNA fragmentomic features were analyzed and incorporated into machine learning models. The models were evaluated using 10-fold cross-validation and assessed for accuracy, sensitivity, specificity, and AUC values. We also performed differential analysis of key genomic features, such as Alu elements and long terminal repeats (LTRs), between benign and malignant CRC samples. Results: The machine learning algorithm demonstrated robust discriminative performance across all datasets using generalized linear modeling (GLM), achieving AUC values of 0.959 (training set), 0.979 (internal validation cohort), and 0.959 (external validation cohort). Notably, the model exhibited particularly strong classification accuracy for advanced-stage colorectal cancer (CRC). Comparative cfDNA profiling revealed distinct molecular signatures between benign and malignant samples: benign samples were characterized by elevated frequencies of Alu elements and long terminal repeats (LTRs), whereas malignant samples showed distinct end motif profiles, characterized by the significant enrichment of specific 4-mer end motifs. These findings suggest that these molecular features may serve as potential biomarkers for malignancy detection. Conclusion: This study demonstrates that cfDNA fragmentomic profiling, particularly differential patterns of Alu and LTR elements, effectively discriminates benign from malignant colorectal lesions. These findings validate the clinical utility of repetitive element analysis and provide a foundation for developing improved non-invasive CRC diagnostics through machine learning approaches incorporating genomic features.