Cotton duplicated genes produced by polyploidy show significantly elevated and unbalanced evolutionary rates, overwhelmingly perturbing gene tree topology

A phylogenetic tree can be used to illustrate the evolutionary relationship between a group of genes, especially duplicated genes, which are sources of genetic innovation and often the hotspot of researches. However, duplicated genes may have complex phylogenetic topologies due to changes in their evolutionary rates. Here, by constructing phylogenetic trees using different methods, we evaluated the phylogenetic relationships of duplicated genes produced by polyploidization in cotton. We found that at least 83.2% of phylogenetic trees did not conform the expected topology. Moreover, cotton homologous gene copy number has little effect on the topology of duplicated genes. Compared with their cacao orthologs, elevated evolutionary rates of cotton genes are responsible of distorted tree topology. Furthermore, as to both branch and site models, we inferred likely positive natural selection during the divergence of fibre-development-related MYB genes, and found that the reconstructed tree topology may often overestimate natural selection, as compared to the inference with the expected trees. Therefore, we emphasize the importance to borrow precious information from gene colinearity in tree construction and evaluation, and have evidence to alert the citation of thousands of previous reports of adaptivity and functional innovation of duplicated genes.