Arrow plot for selecting genes in a microarray experiment: An explorative study

Abstract

Genetic expression analysis is essential for the identification of gene functions and take even more importance when they are directly related with diseases. For the performing of a large-scale study of changes in gene expression it is necessary to find a method to do it with precision and accuracy. Thus, the analysis by the microarray technology is an important tool in the diagnosis of diseases. An important role of the analysis of microarray data involves the determination of which genes could be differentially expressed (DE) across two or more kind of tissue samples. The traditional methods to detect DE genes are generally based on simple measures of distances and could failed in this classification. In this work it is explored a new tool proposed by Silva-Fortes [21] that overcome this difficulty, the arrow plot. This tool is also compared with other methods mostly used to this purpose. The arrow plot is a graphical tool based on two measures of distances between two probability density functions: the overlapping coefficient (OVL) between two densities and the area under the receiver operating characteristic (ROC) curve (AUC), for each gene of a microarray experience. For illustrative purpose we will use a dataset of pancreatic adenocar-cinoma. All computation will be done in R software.