The Most Frequent Gene Tree Based on Phylogenomic Data Reveals Insights into the Origins of Land Vertebrates

Whole-genome data provide all genetic information and offer a promising opportunity to reconstruct a species tree. One of the remaining major challenges is to identify a suitable method for reconstructing a species tree from whole-genome data. However, incongruent evolutionary relationships among land vertebrates, lungfishes, and coelacanths are inferred using concatenation or coalescent-based methods with the previous five genome-scale datasets. To address the controversy that has persisted for decades, we employ the most frequent gene tree approach to estimate a species tree from individual gene trees with the same previous five genome-scale datasets. Our findings consistently support lungfishes as the closest living relatives of land vertebrates, showing statistically significant differences in the first and second most frequencies of individual gene trees when compared with the five previous datasets derived from lungfish transcriptomes. Further analyses demonstrate that the top three most frequent trees align with the three hypotheses regarding the phylogenies of land vertebrates, lungfishes, and coelacanths. These insights are not arbitrary conjectures but are supported by the three largest numbers of genes. The most frequent gene tree approach may be a good alternative phylogenetic method for phylogenomic data. The most-to-second ratio is a valid measure of phylogenetic support for phylogenomic data to express the level of the phylogenetic signals.