Figure 3.

Using the MADMuscle meta-analysis tool to perform systematic cross-species and cross-platform comparisons independently from the studied samples. A- For each of the two input gene lists (clusters 1 and 5 presented in Figure 1), the 50 top overlapping clusters from the database were identified. Red and green backgrounds indicate significant overlaps with cluster 1 (DMD+) and cluster 5 (DMD-) respectively. Non-significant overlaps are displayed by a brown background. The 100 best-hit clusters, showing significant similarity with either cluster 1 or 5, cover 5 distinct species and 18 distinct microarray platforms (see color legend). B- Among the 100 top similar clusters, 3 particular results of overlapping genes are illustrated for cluster 1 (I, II and III) and 3 others for cluster 5 (IV, V, VI). For instance, 283 genes had increased expression (similarity p-value p = 1.57e-66) in both DMD muscle (input cluster 1) and mdx muscle (output cluster II). On the contrary, expression of 119 genes was commonly decreased (p = 6.96e-21) in both DMD (input cluster 5) and mdx (output cluster VI) muscles. C- Functional annotation of clusters 1 and 5 was inferred from their 50 best hits. For each of the 100 top clusters, the five top gene ontology (GO) terms are displayed using the same color code as depicted in part A. Cluster 1 resembles clusters enriched in genes coding for proteins of the extracellular matrix and proteins involved in immune response. Cluster 5 has similarity with clusters preferentially containing genes coding for muscle and mitochondrion proteins.

Baron et al. BMC Genomics 2011 12:113   doi:10.1186/1471-2164-12-113
Download authors' original image