Figure 8.

FLN-based disease-disease association network. (a) An example network of 66 diseases (nodes) linked by the top 100 edges with the highest mutual predictability scores. Many of these edges cannot be identified using the simple disease-gene sharing method, and these are colored red. The rest of the edges are colored blue, which link disease pairs with one or more overlapping disease genes. This disease-disease association network has a modular topology, and diseases tend to form clusters (rectangles). The disease clusters are labeled based on enriched disease class or underlying molecular mechanisms. The following disease nodes mentioned in the text are labeled: ovarian cancer (OC), prostate cancer (PC), Leber congenital amaurosis (LC), cone or cone-rod dystrophy (CD), retinitis pigmentosa (RP), night blindness (NB), Leigh syndrome (LS), mitochondrial DNA depletion syndrome (MD), combined oxidative phosphorylation deficiency (CO), mitochondrial complex I/II/III deficiency (MC), Alzheimer's disease (AZ), hypercholesterolemia (HC), pseudohypoaldosteronism (PH), Bartter syndrome (BS), holoprosencephaly (HP), Waardenburg syndrome (WS), central hypoventilation syndrome (CH), Hirschsprung disease (HD), ceroid-lipofuscinosis (CL), and glaucoma (GC). (b) Functional links among Alzheimer's disease genes (purple nodes) and hypercholesterolemia disease genes (blue nodes) in the FLN. Green edges represent functional links between Alzheimer's disease genes and hypercholesterolemia genes. Light blue edges represent functional links connecting genes associated with the same disease. Edge thickness is proportional to linkage weight. Both networks are generated by Visant [76].

Linghu et al. Genome Biology 2009 10:R91   doi:10.1186/gb-2009-10-9-r91
Download authors' original image