Purpose. DNA methylation alterations play important roles in initiation and progression of clear cell renal cell carcinoma (ccRCC). In this study, we attempted to identify differentially methylated mRNA signatures with prognostic value for ccRCC. Methods. The mRNA methylation and expression profiling data of 306 ccRCC tumors were downloaded from The Cancer Genome Atlas (TCGA) to screen differentially methylated lncRNAs and mRNAs (DMLs and DMMs) between bad and good prognosis patients. Uni- and multivariable Cox regression analyses and LASSO Cox-PH regression analysis were used to select prognostic lncRNAs and mRNAs. Corresponding risk scores were calculated and compared for predictive performance in the training set using Kaplan-Meier OS and ROC curve analyses. The optimal risk score was then identified and validated in the validation set. Function enrichment analysis was conducted. Results. This study screened 461 DMMs and 63 DMLs between good prognosis and bad prognosis patients, and furthermore, nine mRNAs and six lncRNAs were identified as potential prognostic molecules. Compared to nine-mRNA status risk score model, six-lncRNA methylation risk score model, and six-lncRNA status risk score model, the nine-mRNA methylation risk score model showed superiority for prognosis stratification of ccRCC patients in the training set. The prognostic ability of the nine-mRNA methylation risk score model was validated in the validation set. The nine prognostic mRNAs were functionally associated with neuroactive ligand receptor interaction and inflammation-related pathways. Conclusion. The nine-mRNA methylation signature (DMRTA2, DRGX, FAM167A, FGGY, FOXI2, KRTAP2-1, TCTEX1D1, TTBK1, and UBE2QL1) may be a useful prognostic biomarker and tool for ccRCC patients. The present results would be helpful to elucidate the possible pathogenesis of ccRCC.

中文翻译：

---

基于九种差异甲基化 mRNA 的风险评分模型用于预测透明细胞肾细胞癌患者的预后

目的。DNA甲基化改变在透明细胞肾细胞癌（ccRCC）的发生和发展中起重要作用。在这项研究中，我们试图识别具有 ccRCC 预后价值的差异甲基化 mRNA 特征。方法. 从癌症基因组图谱 (TCGA) 下载了 306 个 ccRCC 肿瘤的 mRNA 甲基化和表达谱数据，以筛选预后不良和良好预后患者之间差异甲基化的 lncRNA 和 mRNA（DML 和 DMM）。单变量和多变量 Cox 回归分析和 LASSO Cox-PH 回归分析用于选择预后 lncRNA 和 mRNA。使用 Kaplan-Meier OS 和 ROC 曲线分析计算相应的风险评分并比较训练集中的预测性能。然后在验证集中确定和验证最佳风险评分。进行了功能富集分析。结果. 本研究筛选了预后良好和预后不良患者之间的 461 个 DMM 和 63 个 DML，此外，9 个 mRNA 和 6 个 lncRNA 被鉴定为潜在的预后分子。与 9-mRNA 状态风险评分模型、6-lncRNA 甲基化风险评分模型和 6-lncRNA 状态风险评分模型相比，9-mRNA 甲基化风险评分模型在训练集中 ccRCC 患者的预后分层方面显示出优势。在验证集中验证了九个 mRNA 甲基化风险评分模型的预后能力。九种预后 mRNA 在功能上与神经活性配体受体相互作用和炎症相关通路相关。结论. 九个 mRNA 甲基化特征（DMRTA2、DRGX、FAM167A、FGGY、FOXI2、KRTAP2-1、TCTEX1D1、TTBK1 和 UBE2QL1）可能是 ccRCC 患者有用的预后生物标志物和工具。目前的结果将有助于阐明ccRCC的可能发病机制。