Figure 3

Rooted phylogenetic tree based on the pol amino acid sequences of the DIRS1-like families identified. Distances are calculated with JTT parameter model plus gamma distribution's correction for amino acids. The tree is constructed using the Neighbor Joining method and pairwise deletion of gaps option included in MEGA5.0 software. When possible, one representative copy sequence that required only minor corrections for each family was integrated into our analysis. Reference elements are labeled with an asterisk and clusters that correspond to an element annotated in this study are written in bold italics. If a reference element was included in a family, this sequence was chosen to represent the family. In the cases of Anolis carolinensis, Daphnia pulex and Xenopus tropicalis, species that show a high family number, only four or five of their most abundant families were integrated. Ty3/Gypsy element sequences were used as outgroups according to the close relationships of their reverse transcriptase and RNase H domains with those of DIRS1-like and PAT retrotransposons. Support for individual groups was evaluated with non-parametric bootstrapping using 100 replicates. Only bootstrap node values over 50% are represented.