Figure 7

Hierarchical cluster of a pairwise similarity matrix and ontological analysis to identify relatedness between animal models and human cardiac disease. A similarity matrix was produced for each model or human disease type, with each pairwise comparison score calculated as the number of genes altered in the same direction (up-regulated or down-regulated at least 1.5-fold on average) compared to the appropriate study controls. The resulting scores were log transformed and Z scores calculated before performing hierarchical clustering using CLUSFAVOR 6.0 software (A). Bright red represents the highest pairwise similarity scores, black represents median values, and bright green represents the lowest pairwise similarity values. Each gene list was also analyzed using Ingenuity Pathway Analysis software and function categories for each animal model and human heart failure type compared (B). Functional categories shown were significantly over-represented (B-H multiple hypothesis correction p value ≤ 0.05), and numbers represent the number of altered genes in that category.