作者: G.-H. Zhang , Q.-Y. Zhong , X.-X. Gou , E.-X. Fan , Y. Shuai
DOI: 10.1007/S12094-019-02057-3
关键词:
摘要: Glioma is a common malignant tumor of the central nervous system, which characterized by low cure rate, high morbidity, and recurrence rate. Consequently, it imperative to explore some indicators for prognostic prediction in glioma. We obtained glioma data from The Cancer Genome Atlas (TCGA). Differentially expressed genes (DEGs) were R software TCGA sets. Through Cox regression analysis, risk scores assess weighted gene-expression levels, could predict prognosis patients with validity value this model confirmed manifestation receiver-operating characteristic (ROC) curves, area under curve (AUC), 5-year overall survival (OS). In total, 920 DEGs transcriptome extracted database. identified novel seven-gene signature associated Among them, AL118505.1 SMOC1 positively related OS glioma, showing better glioma; however, RAB42, SHOX2, IGFBP2, HIST1H3G, IGF2BP3 negatively OS, displaying worse prognosis. addition, according scores, was also protective factor, while others factors. Furthermore, expression levels significantly correlated grades. Receiver-operating assessed accuracy sensitivity gene signature. Each seven distribution score presented heat map. be used as predictor future.