Distribution of positively selected sites.

Distribution of the number of positively selected amino acid sites after correction of p-values with a false discovery method for the data set containing three rust genomes (basidioPAML) (a) and the data set containing the same rust genomes in addition to EST data from Hemileia vastatrix (basidioPAML_Hv) (b). Embedded in each histogram is a doughnut chart with the distribution of the positively selected sites across the two main site class pairs defined in this study for the basidioPAML data set (a) and basidioPAML_Hv data set (b). Unique sites represent amino acids exclusive and identical in all rust species and Diversifying sites represent amino acids exclusive but variable in rust species. The site classes are colour coded with the corresponding legend on the bottom of the figure.