Figure 6

Co-functional modules of direct LXR target genes. Associations of LXR target genes with the annotations from Reactome, CGAP tissue EST expression, KEGG and GO databases clustered and visualized using heatmap.2 function from R-package gplots [38]. The y-axis columns indicate the annotations and x-axis rows the associated genes. Each association, depicted as a cell of the heat map, has been weighted using the multiplication of log₂ FC of a gene (row) and -log₁₀ P-value of annotation (column) enrichment highlighting the most important associations. Red and green color scales are used for up- and down-regulated genes, respectively. Both columns and rows have been clustered using agglomerative hierarchical clustering with asymmetric binary distance measure. The eight gene and annotation clusters cover the majority of the gene set. Indicated is also the LXR peak strength density graph, which summarizes the peak heights over the genes on the x-axis. For each gene the peak strength has been calculated as a sum of the log₂ FEs for the highest peak from the close (± 100 kb) and distant region (± 0.1 to ± 1 Mb).