Neighborhood analysis. Panel A depicts 2 samples as vectors in gene expression space. The coordinates of the sample vector are composed of expression levels (g₁, g₂,..., g_n) for each gene in the sample. The distance (d) can be calculated between 2 samples. Panel B is a schematic of a neighborhood analysis. A correlation is calculated between the expression of gene S and other genes with similar patterns of expression across different samples. On the right side of panel B, the coordinates of the genes have been randomly permuted, so the number of genes that correlate with S′ at a given level of significance is decreased. Panel C graphically illustrates a neighborhood analysis. The number of genes in a neighborhood increases as the measure of correlation decreases.