Visualization of gene expression data along with information from various genome databases is the most effective way of deciphering the statistical relationships within the data samples. We typically use 1) scatter plot to examime the relationship between two samples, 2) Color-mosaic images to represent the expression ratios from different experiments, and 3) 2D or 3D multidimensional scaling plot to examine the similarity among all samples, or the dendrogram to sort the order of samples or genes.

Multidimensional Scaling (MDS). Multidimensional scaling is a method that represents measurements of similarity (between pairs of objects), as Euclidean distance between points in a dimension-reduced space. Typically, we use the distance matrix, as illustrated in Table 1, in the clustering section, as a similarity matrix between all pairs of samples in one experiment design. The dimension reduction procedure is to minimize the difference between the distance in higher-dimensional space and the approximate distance in the dimension-reduced (2D or 3D) space. Table 1 can be represented by the figure shown below.

Figure 1. MDS plot of distance table 1 in Clustering Section.