Maximum-likelihood phylogenetic tree of <i>Trichosporon</i> species, based on analysis of the ITS1 and ITS2 regions and the D1/D2 region of the LSU.

<p>Strain names appear after the species name. <i>Trichosporon</i> species of clinical significance appear in bold. See <a href="http://www.plospathogens.org/article/info:doi/10.1371/journal.ppat.1004982#ppat.1004982.s001" target="_blank">S1 Table</a> for the GenBank accession numbers. <i>Cryptococcus neoformans</i> was used as an outgroup. Sequences of each individual region were structurally aligned using MXSCARNA [<a href="http://www.plospathogens.org/article/info:doi/10.1371/journal.ppat.1004982#ppat.1004982.ref058" target="_blank">58</a>] and then concatenated into a supermatrix using FASConCAT [<a href="http://www.plospathogens.org/article/info:doi/10.1371/journal.ppat.1004982#ppat.1004982.ref059" target="_blank">59</a>]. Phylogenetic inference was carried out with the software RAxML [<a href="http://www.plospathogens.org/article/info:doi/10.1371/journal.ppat.1004982#ppat.1004982.ref060" target="_blank">60</a>]. For the loop regions, the evolutionary model GTR was used. The stem regions were analyzed under the S16 evolutionary model, thus taking into account secondary structure topology, i.e., compensatory mutations [<a href="http://www.plospathogens.org/article/info:doi/10.1371/journal.ppat.1004982#ppat.1004982.ref061" target="_blank">61</a>–<a href="http://www.plospathogens.org/article/info:doi/10.1371/journal.ppat.1004982#ppat.1004982.ref063" target="_blank">63</a>]. Substitution rate heterogeneity was taken into account using the gamma model of Yang [<a href="http://www.plospathogens.org/article/info:doi/10.1371/journal.ppat.1004982#ppat.1004982.ref064" target="_blank">64</a>]. Bootstrap values were computed over 1,000 replicates. Grey dots represent bootstrap values 50%–75% and black dot bootstrap values 76%–100%. Bar, 0.07 substitutions per nucleotide position. T: Type strain.