Fig. 4

Node-graph visualization of the GO terms describing GP1 and GP2 genes, as determined by the REVIGO algorithm [24]. Larger dots represent more relevant GO terms, characterized by relative enrichment. Note that GP1 and GP2 gene ontology categories are linked within each of the two groups by a higher number of interconnections than across the two groups. This suggests that within GP1 and GP2 there are similarities. Also note that, as detailed in the main test, upon randomization of the genes assigned to the two groups, there is a strong (~ 26 fold) increase of number of interconnections between the two groups, implying that the separation of the two groups of genes is not stochastic. For a list of these categories, see Additional file 1