TCGA High Grade Serous Ovarian Cancer Features for Subtype Identification

Survival-associated features of mRNA expression, microRNA expression, copy number alteration (CNA), and DNA methylation from 512 TCGA HGSOC samples (TCGA-HGSOC) from Zhang et al (2013) (Zhang, W., et al., Integrating genomic, epigenomic, and transcriptomic features reveals modular signatures underlying poor prognosis in ovarian cancer. Cell Rep, 2013. 4(3): p. 542-53.). The prognostic significances of individual features were assessed by univariate Cox proportional hazard model analysis (p &lt; 0.05) relative to patient survival. A total of 4,526 features are included, including the expression of 1,651 mRNA genes, 140 microRNAs, 2,191 somatic CNAs, and 455 DNA methylation sites. Feature values were then normalized with respect to the expression of normal controls as previously described by Zhang et al (2013).本数据集包含来自Zhang等（2013年）的512例TCGA高级别浆液性卵巢癌（TCGA-HGSOC）样本的mRNA表达、微小RNA（microRNA）表达、拷贝数变异（CNA）及DNA甲基化相关生存特征[Zhang W, et al. Integrating genomic, epigenomic, and transcriptomic features reveals modular signatures underlying poor prognosis in ovarian cancer. Cell Rep, 2013, 4(3): 542-53.]。针对患者生存情况，采用单变量Cox比例风险模型评估单个特征的预后意义，显著性阈值设定为p < 0.05。本数据集共计纳入4526个特征，其中包括1651个mRNA基因表达特征、140个microRNA表达特征、2191个体细胞CNA特征以及455个DNA甲基化位点。随后，参照Zhang等（2013年）已报道的方法，以正常对照的表达水平为基准对所有特征值进行归一化处理。