Extended Data Fig. 5: Phylogenetic analyses of Laurasiatheria.

a, A total of 10,857 conserved noncoding elements (CNEs) were used to determine a mammalian phylogeny using noncoding regions (topology 2 in Supplementary Fig. 4). Bootstrap support values less than 100 are displayed, with internal nodes that differ to the protein-coding supermatrix highlighted in red. The position of Chiroptera as basal to Fereuungulata, as in Fig. 2, is maintained. b, All gene alignments were fit to the 15 laurasiatherian topologies (Supplementary Fig. 4) we explored, to determine which tree had the highest likelihood score for each gene. The number of genes supporting each topology is displayed. c, A supermatrix consisting of 1st and 2nd codon sites from 448 genes that are evolving under homogenous conditions—thus considered optimal ‘fit’ for phylogenetic analysis—was used to infer a phylogeny using maximum likelihood (topology 13 in Supplementary Fig. 4). Bootstrap support values less than 100 are displayed, with internal nodes that differ to the protein-coding supermatrix phylogeny highlighted in red. Unlike Fig. 2, Chiroptera is now sister to (Carnivora + Pholidota); however, this split has low bootstrap support (58%). d, Using the 488 genes considered fit for phylogenetic analyses, the position of bats within Laurasiatheria under a model of coalescence using SVDquartets. The resulting phylogeny is displayed. The tree is rooted on Atlantogenata, with support values from bootstrap pseudoreplicates. Only nodes with support less than 100 have their values displayed. The position of Chiroptera as basal to Fereuungulata, as in Fig. 2, is maintained.