可调FPR下利用基因芯片实验数据筛选差异表达基因

Screen of significant genes with adjustable FPR based on microarray data

在传统t检验和方差分析的基础上,通过对重复实验数据的随机组合,提出了一种计算可调节假阳性率(FPR)的方法,并将之应用到基因芯片表达数据的分析中,目的是筛选差异表达基因.通过在模拟数据集和真实基因芯片表达数据的应用,均取得较好的效果.认为该方法适用于基因芯片表达数据,进行差异表达基因的筛选,并且比传统的方法更具灵活性.

Based on conventional t test and ANOVA, a method was brought up to screen significant genes based on microarray data. With the method, replicated gene expression data was random combined and adjustable FPR(False Positive Rate) was computed. Idealized operation result was made with the method used in simulative data and real microarray data. It was concluded that the method was suitable to analyze gene expression data and screen significant genes. Besides, the method was more flexible than conventional methods.