Figure 4

Phylogeny of Runx (A) and CBFβ (B) proteins. In both cases, maximum likelihood trees were constructed using the program PhyML. In the case of Runx, likelihoods were calculated using the Rt-REV amino acid substitution matrix [83]. In the case of CBFβ, likelihoods were calculated using the JTT amino acid substitution matrix. Separate neighbor-joining analyses using the JTT distance matrix were also performed for each data set (not shown). The bars and icons indicate the locations of the taxonomic groupings on each tree with the deuterostome and protostome groups marked with filled and empty circles respectively. Both trees are rooted between the sponge sequences and the eumetazoan sequences. The length of each horizontal branch is proportional to the number of amino acid substitutions that have occurred along that branch (scale at lower part of each panel). The first and second numbers at each node indicate the percentage of 1,000 bootstrap replicates in which the given clade is recovered with maximum likelihood and neighbor-joining analyses, respectively. Nodes which failed to receive at least 40% bootstrap support in one of the analyses are not labeled. The sources of the sequences and the abbreviations for taxa are provided in Additional files 9 and 10.