General features of the three genomes.

Data are based on the manual annotation of MED134, GenDB annotation in the case of PRO95, and 4H-3-7-5 GenBank file. Peptidases, glycosyl hydrolases, adhesion proteins and transporter genes were recalculated as the number of hits to the corresponding PFAMs, and the genes in paralogous clusters were calculated as described [28]. Lipoproteins were identified using LipoP (http://www.ncbi.nlm.nih.gov/pubmed/12876315) and signal peptides with SignalP (http://www.ncbi.nlm.nih.gov/pubmed/21959131).