Phylogenomic Resolution of the Phylogeny of Laurasiatherian Mammals: Exploring Phylogenetic Signals within Coding and Noncoding Sequences

Abstract

The interordinal relationships of Laurasiatherian mammals are currently one of the most controversial questions in mammalian phylogenetics. Previous studies mainly relied on coding sequences (CDS) and seldom used noncoding sequences. Here, by data mining public genome data, we compiled an intron data set of 3,638 genes (all introns from a protein-coding gene are considered as a gene) (19,055,073 bp) and a CDS data set of 10,259 genes (20,994,285 bp), covering all major lineages of Laurasiatheria (except Pholidota). We found that the intron data contained stronger and more congruent phylogenetic signals than the CDS data. In agreement with this observation, concatenation and species-tree analyses of the intron data set yielded well-resolved and identical phylogenies, whereas the CDS data set produced weakly supported and incongruent results. Further analyses showed that the phylogeny inferred from the intron data is highly robust to data subsampling and change in outgroup, but the CDS data produced unstable results under the same conditions. Interestingly, gene tree statistical results showed that the most frequently observed gene tree topologies for the CDS and intron data are identical, suggesting that the major phylogenetic signal within the CDS data is actually congruent with that within the intron data. Our final result of Laurasiatheria phylogeny is (Eulipotyphla,((Chiroptera, Perissodactyla),(Carnivora, Cetartiodactyla))), favoring a close relationship between Chiroptera and Perissodactyla. Our study 1) provides a well-supported phylogenetic framework for Laurasiatheria, representing a step towards ending the long-standing "hard" polytomy and 2) argues that intron within genome data is a promising data resource for resolving rapid radiation events across the tree of life.

Keywords: Laurasiatheria; data subsampling; intron; noncoding; phylogenomics; phylogeny.

Fig. 1.—

Characteristics of CDS and intron data sets (blue = CDS, red = intron). Boxplots show (A) variation in gene length, (B) GC content of each gene (among genes) and each species (among species), (C) relative evolutionary rates of loci (measured by the average pairwise distance for each gene), and (D) average bootstrap support values across all estimated gene trees. (E) Visualization of ML tree space using multidimensional scaling plot of 10,259 ML gene-trees from the CDS data set; each dot represents a tree inferred from one gene. Distances between dots represent Robinson–Foulds distances between gene trees. (F) Multidimensional scaling plot of 3,638 ML gene-trees from the intron data set. (G) Histogram of the average RF distance for a gene relative to all other genes, summarized from the CDS data set and the intron data set.

Fig. 2.—

Phylogenetic relationships of Laurasiatheria inferred from the CDS data set (10,259 genes; 20,994,285 sites). Phylogeny was inferred by concatenation ML and species tree analysis using the ASTRAL program. The ML phylogeny is shown on the left, and the ASTRAL species tree is shown on the right (outgroup not shown). Values next to branches are bootstrap values. Branches without support values all received a bootstrap value of 100%.

Fig. 3.—

Phylogenetic relationships of Laurasiatheria inferred from the Intron data set (3,638 genes; 19,055,073 sites). Phylogeny was inferred by concatenation ML and species tree analysis using the ASTRAL program. Both analyses produced identical phylogenies for the interordinal relationships of Laurasiatherian mammals. All branches have a bootstrap value of 100% in both analyses. Branch lengths are from the ML analysis.

Fig. 4.—

Phylogenetic inference robustness for the CDS and Intron data sets, which were resampled into 11 data subsets under different data subsampling criteria (see Materials and Methods for details). These data subsets were analyzed with both concatenated and species-tree inferences. (A) There are in total seven topologies found in these phylogenetic analyses (each color represents a specific tree topology). Bootstrap support for certain topologies from different data subsets are shown in charts (B) through (E).

Fig. 5.—

Effect of outgroup choices on phylogenetic inferences of the CDS data set (A) and the Intron data set (B). There are three outgroup combination schemes: “X + E” refers to the use of Xenarthra and Euarchontoglires as outgroup; “A + E” refers to the use of Afrotheria and Euarchontoglires as outgroup; “A + X” refers to the use of Afrotheria and Xenarthra as outgroup. Note that the change in outgroup has no effect on the phylogenetic inference of the Intron data set but can influence the phylogenetic inference of the CDS data set. Chiro: Chiroptera, Cetartio: Cetartiodactyla, Perisso: Perissodactyla, Carni: Carnivora.

Fig. 6.—

Gene-support frequency statistics for the 15 alternative hypotheses (H1–H15) regarding the interrelationships of Chiroptera, Perissodactyla, Carnivora, and Cetartiodactyla. Intron data are displayed in blue, and CDS data are displayed in green. Genes whose gene trees do not support any of the 15 alternative hypotheses are considered “nonmatching.” Gene-tree statistics are based on “matching” genes only. The histograms on the left show the proportion of gene trees that support a given hypothesis.