Figure 2.

The Visualization Experts' pick. (a) Association gene network ed from all pairs of 3843 SNPs with a significant association (p <0.05, PLINK two-locus results) with the gene expression of the 15 genes and filtered such that only SNP pairs containing at least one highly significant SNP (R2 >0.1 and p <0.05, PLINK single locus results) remain. All edges with weight w ≥ 40 are shown. Nodes represent genes, edges represent significant SNP pairs. Genes significantly associated with SNP pairs are colored using a distinct color, genes with no significant association are drawn with gray fill. Each edge conveys four pieces of information: An edge e of weight w starting in node s, ending in node t and drawn with color c represents w SNP pairs, where each of them has one SNP in gene s and one in gene t. These SNP pairs are significantly associated with the expression of the gene whose node is filled with color c; (b) Aggregated iHAT visualization of 29 visually selected SNPs where the 'affected' and 'not affected' groups display different colors.

Bartlett et al. BMC Bioinformatics 2012 13(Suppl 8):S8   doi:10.1186/1471-2105-13-S8-S8