Systematic Analysis of 4-gene Prognostic Signature in Patients with Diffuse Gliomas Based on Gene Expression Profiles

Background: Diffuse gliomas are a group of diseases that contain different degrees malignancy and complex heterogeneity. Previous studies proposed biomarkers for certain grades gliomas, but few them have conducted systematic analysis to search molecular markers. Methods: WGCNA was used find significant genes associated with malignant progression diffuse glioma in TCGA sequencing expression data the GEO profile-merge meta dataset. Lasso regression potential model building best selected by CPE, IDI, C_index. Risk score evaluate gene signature prognostic power. Multi-omics data, including CNV, methylation, clinical traits, mutation, were evaluation. Results: We found out 67 significantly WGCNA. Next, we established new 4 marker (KDELR2, EMP3, TIMP1, TAGLN2). Multivariate cox identified risk as an independent predictor prognosis patients its predictive power histopathological glioma. Further, had confirmed five test datasets remained good The combination specific characteristics possessed better Furthermore, divided low-risk into three subtypes: LowRisk_IDH1wt, LowRisk_IDH1mut/ATRXmut, LowRisk_IDH1mut/ATRXwt combining IDH1 mutation ATRX which obvious survival difference. In further analysis, multi-omics features. Conclusion: malignant-related 4-gene profile from multiple microarrays data. four markers on overall gliomas' genetic backgrounds, features, signal pathways.