Table 3.
Summary of Data Quality and Phylogenetic Tests Applied to 66TaxonSet and 39TaxonSet.
| Data Sets | Compositional Heterogeneity Detected | P4 | PhyloBayes | Root Position (P4/PhyloBayes) | Reject Competing Hypotheses 2Ln(BF) = 6–10 | Single strongly Supported Hypothesis |
|---|---|---|---|---|---|---|
| 66TaxonSet_nuc | Yes | 3GTR + 5C + I + 4Γ | 1GTR + 1C + 4Γ | Atlantogenataa/Atlantogenata | Afrotheria and Xenarthra | None |
| 66TaxonSet_aa | Yes | 1JTT + 4C + I + 4Γ | CAT | Xenarthraa/Atlantogenata | Afrotheria and Atlantogenata | None |
| 66TaxonSet_day | No | 5GTR + 2C + I + 4Γ | 1GTR + 1C + 4Γ | Xenarthra/Xenarthraa | None | None |
| 39TaxonSet_nuc | Yes | 2GTR + 4C + 4Γ | CAT | Atlantogenata/Atlantogenataa | None | None |
| 39TaxonSet_aa | Yes | 1JTT + 5C + 4Γb | CAT_GTR | Atlantogenata/Atlantogenata | — | Atlantogenata |
| 39TaxonSet _day | Yes | 2GTR + 4C + 4Γ | CAT_GTR | Atlantogenata/Atlantogenata | Afrotheria and Xenarthra | Atlantogenata |
Note.—The results of the χ2 test of compositional homogeneity are shown. The model of best fit from P4 and PhyloBayes analyses are given. The placental root supported by each data set is indicated for both the P4 and the PhyloBayes analyses. The ability of each data set to reject alternative rooting hypotheses is given (2(ln BF) > 6 indicates alternative rooting hypotheses can be strongly rejected).
aRooting hypotheses that have a PP support value of <0.50, that is, poor support. A hypothesis is denoted as being the “single strongly supported hypothesis” if it has both 1) a high PP at the placental root and 2) all other hypotheses could be rejected.
bThis model did not fit the data.