A metaanalysis of bat phylogenetics and positive selection based on genomes and transcriptomes from 18 species

Hawkins, John A.; Kaczmarek, Maria E.; Müller, Marcel A.; Drosten, Christian; Press, William H.; Sawyer, Sara L.

Understanding the evolutionary history of bats is important not just for the study of Chiropteran zoology but also for the study of bats as reservoirs of deadly human viruses. Knowledge of an accurate phylogeny improves analysis of positive selection in bat genomes because dN/dS analysis requires both gene alignments and a phylogenetic relationship of the orthologs being analyzed. Also, the reliable identification of gene orthologs will allow molecular biologists to functionally test differences in these genes from one species to the next (62) . Functional studies such as these will allow us to understand whether some bats have unique features of their immunity that allow them to harbor viruses that are dangerous to humans (63) . Herein, we have curated multiple sequence alignments of thousands of bat genes. Using both genomic and transcriptomic data, we were able to find 11,677 orthologous gene families. To enhance these alignments, we provided transcriptome data for two of these bats, H. monstrosus and R. aegyptiacus, from which we annotated 7,858 and 9,682 genes, respectively. We furthermore developed a general data cleaning method for filtering exons with nonrandom structural errors, in this case observed to result from genomic vs. transcriptomic data. For this method, we developed the MIXR software package which directly detects and removes alternate consensus run artifacts, and is available at http://github.com/hawkjo/mixr. The multiple sequence alignments that we have created for bat genes, both before and after exon filtering, are available for use by the wider bat and virology communities (http://numerical.recipes/chiroptera/). Using these alignments, we examined the history both of speciation and of positive selection in 18 species of bats. This study hopefully sets the stage for continued and more in-depth study of the evolution and functional differentiation of bat genes relevant to immunity and beyond. Using these orthologous gene families, we were able to reconstruct the phylogeny of the order Chiroptera using multiple methods. Due to the sheer scale of the data, we resolved each node in the tree with 100% reported posterior probability, although the topology differed slightly depending on the analysis method. Our results support the division of Chiroptera into the two suborders Yinpterochiroptera and Yangochiroptera, in disagreement with the traditional division into Megachiroptera and Microchiroptera. However, we acknowledge that rooting ancient clades continues to be a difficult phylogenetic problem, and further data may shed more light on this issue. We furthermore provide evidence for the placement of M. schreibersii, in which we agree with Hoofer and Bussche, supporting their proposal for the separation of Miniopteridae into its own family (53) . We also provide evidence for the disruption of proposed subfamily Phyllostominae by D. rotundus. Most intriguingly, we saw M. leucogaster placed in the Myotis genus, which will require further investigation.