Statistical theory and the evaluation of Cox's proportional hazards model regularized by various penalties for survival DNA microarray expression data

在本研究中，我们使用R语言编写程序，计算各种惩罚正则化的Cox比例风险模型的统计特性，包括基于预后指数的统计功率，并对不同条件下的DNA微阵列表达数据与生存时间进行模拟实验。结果表明，当候选基因数量大于样本量时，惩罚方法的验证集功率在许多情况下大于逐步方法。这种趋势在包括l1规范和l2规范在内的惩罚方法中最为显著。此外，我们将程序与实际的微阵列基因表达数据和生存时间数据进行验证，以确认其有效性。我们的Cox比例风险模型的模拟程序对规划DNA微阵列研究或评估此类研究的结果非常有用。

In this study, we created programs using the R language to calculate statistical properties of Cox's proportional hazards model regularized by various penalties, including the statistical power based on the prognostic index, and conducted simulation experiments under various conditions of DNA microarray expression data with survival time. The results showed that when the number of candidate genes is larger than the sample size, the power of a validation set for penalized methods is greater than for stepwise methods in many cases. This tendency is most remarkable for the penalized methods including both the L1-norm and the L2-norm. Furthermore, we verified our programs to actual microarray gene expression data with survival time data to confirm their validity. Our simulation programs for Cox's proportional hazards model regularized by various penalties are very useful for planning DNA microarray studies or for evaluating the results of such studies.