Figure 4.

Phylogenomic analyses. (a) Amino-acid sequences from 4,689 genes (for 7 Sauropsida, 3 mammals, and 2 outgroup taxa) analyzed with RaxML (WAG model and approximate rate heterogeneity) after removal of excessively gapped positions (final dataset size = 2,012,759 characters/species), as well as with RaxML (GTR model and approximate rate heterogeneity) and MetaPIGA-2 (GTR, gamma-rate heterogeneity) after removal of all gapped positions (final size of dataset = 1,612 characters per species); labels on nodes indicate bootstrap proportions under RaxML for the 2 million and 1,612 aa datasets as well as posterior probabilities generated by MetaPIGA for the 1,612 aa dataset; branch lengths are indicated for the MetaPIGA analysis. (b) Amino-acid sequences from 4,689 genes (for 5 Sauropsida lineages, including 2 hybrid sequences, 3 mammals, and 1 hybrid outgroup) after removal of all gapped positions (final dataset size = 24,071 characters/species) analyzed with RaxML (GTR, approximate rate heterogeneity) and MetaPIGA (GTR, gamma-rate heterogeneity); bootstrap proportions and posterior probabilities are 100% for all branches; MetaPIGA branch lengths are indicated. (c) Amino-acid sequences from 1,139 genes devoid of known paralogs (for 7 Sauropsida species, 3 mammals and 2 outgroup taxa) analyzed with MetaPIGA after removal of excessively gapped positions (final size of dataset = 246,208 characters/species); labels on nodes indicate posterior probabilities for analyses under GTR with/without gamma rate heterogeneity; analysis of this dataset under RaxML still generated long-branch attraction: (corn snake + bearded dragon) and (crocodile + turtle).