Whole genome protein families were classified by InterproScan analysis (http://www.ebi.ac.uk/interpro/) in combination with the Treefam methodology that defines a protein family as a group of genes descended from a common ancestor [121] (link). To identify potential pathogenicity and virulence genes, whole genome blast searches were conducted against protein sequences in the pathogen-host interaction database (version 3.2, http://www.phi-base.org/) (E<1×10−5). The families of proteases were additionally classified by Blastp against the MEROPS peptidase database (http://merops.sanger.ac.uk/). Transporters were classified based on the Transport Classification Database (http://www.tcdb.org/tcdb/). The cytochrome P450s were named according to Dr. Nelson's P450 database (http://drnelson.utmem.edu/CytochromeP450.html). G-protein coupled receptors, protein kinases, transcription factors and GH families were classified by Blastp against GPCRDB (http://www.gpcr.org/7tm/), KinBase (http://kinase.com/), Fungal Transcription Factor Database (http://ftfd.snu.ac.kr/) and CAZy database (http://www.cazy.org/), respectively. All Metarhizium genes with significant hits (E value ≤ 10−5) to GPCRDB sequences and that contained 7 transmemebrane helices (analyzed with http://www.cbs.dtu.dk/services/TMHMM/) were included as putative GPCRs. To analyze fungal secondary metabolite pathways, the genome annotation data from both species were coordinated and analyzed with the program SMURF (http://www.jcvi.org/smurf/index.php). The evolution of protein family size variation (expansion or contraction) was analyzed using CAFE [32] (link).