Abstract

Discovering biomarkers is central to the research and treatment of degenerative central nervous system (CNS) diseases, playing a crucial role in early diagnosis, disease monitoring, and the development of new treatments, particularly for challenging conditions like degenerative CNS diseases and glioblastoma (GBM). This study analyzed gene expression data from a public database, employing differential expression analyses and Gene Co-expression Network Analysis (WGCNA) to identify gene modules associated with degenerative CNS diseases and GBM. Machine learning methods, including Random Forest, Least Absolute Shrinkage and Selection Operator (LASSO), and Support Vector Machine - Recursive Feature Elimination (SVM-RFE), were used for case-control differentiation, complemented by functional enrichment analysis and external validation of key genes. Ninety-five commonly altered genes related to degenerative CNS diseases and GBM were identified, withandemerging as significant through machine learning screening. Receiver operating characteristic (ROC) analysis confirmed their diagnostic value, which was further validated externally, indicating their elevated expression in controls. The study's integration of WGCNA and machine learning uncoveredandas potential biomarkers for degenerative CNS diseases and GBM, suggesting their utility in diagnostics and as therapeutic targets. This contributes new perspectives on the pathogenesis and treatment of these complex conditions.