【摘要】 为精确预测食管癌病人的存活风险,通过构建食管癌编码基因(mRNA)和非编码基因(ncRNA)共表达网络,利用随机游走挖掘网络节点的拓扑权重,并与随机生存森林(RSF)整合,构建食管癌风险预测模型.结合5倍交叉证实和独立数据集验证,利用C-index评估算法预测性能.识别出了用于构建风险预测模型的最优lnc RNA分子标签,并实现了食管癌生存风险的精确预测.更多还原

【Abstract】 To accurately predict the survival risk of patients with esophageal squamous cell carcinoma(ESCC), constructed a protein coding genes(m RNA)-non-coding genes(nc RNA)network,and evaluated the topological weights of network nodes by random walk.Integrated the topological weights into random survival forest(RSF),and constructed a risk prediction model of ESCC.The performance of the algorithm was evaluated based on five-fold cross-validation and independent dataset verification.The optimal lnc RNA molecular signature used to construct the risk prediction model was identified and the accurate prediction of the survival risk of esophageal cancer was realized.更多还原