Peptides for all 8 species were obtained. All but the longest isoform of each gene was removed and an all-v-all BLAST search was performed. The e-values from the BLAST search serve as the main clustering criterion for MCL, which groups peptides into gene families. These gene families, along with the species phylogeny (obtained from ML estimation of gene trees and based on dates from Timetree.org) were given as input to CAFE to estimate gene gain and loss rates (lambda) and reconstruct ancestral gene counts..