Fig. 2

Phylogenetic inference from 47 concatenated proteins. a Location of selected species in the previously inferred SSU rDNA phylogeny [10]. b–f Phylogenetic trees inferred by Bayesian inference (b–d), RAxML (e) or Phylobayes (f) from an alignment of 47 concatenated orthologous proteins that were identified in the species shown in a. Bayesian analyses were run for 100,000 generations with either a mixed amino-acid model (b,c) for the entire alignment, or with a partitioned alignment in which each protein was run under its optimal amino-acid substitution model (d). All analyses converged within 6000 generations (SD of split frequencies = 0). The RaxML analysis was run with 100 bootstrap replicates on an alignment partitioned as in d. Phylobayes MPI [22] was run over two chains under the CAT-GTR model (f). Trees were rooted using A. castellani as outgroup. The average GC content of the genomic DNAs encoding the 47 proteins is plotted onto the phylogeny in panel c. Posterior probabilities or bootstrap support for the nodes are shown. Abbreviated and full species names are colour-coded to reflect the taxon group to which the species belong