Figure 5

Example of CPP computation. a) A hierarchical clustering is performed for a given algorithm and the clusters are selected to characterize the original co-association of the genes (with j = 4 in the example). b) After the introduction of the MVs, clusters are selected to characterize the new co-association of the genes. c) The number of common genes is computed between the reference clustering (ref) and the generated clustering (new). In yellow is given the maximum number of common genes allowing the identification of the closest cluster for each cluster. d) The corresponding CPP value is equal to 76.9%