Tackling Rapid Radiations With Targeted Sequencing
In phylogenetic studies across angiosperms, at various taxonomic levels, polytomies have persisted despite efforts to resolve them by increasing sampling of taxa and loci. The large amount of genomic data now available and statistical tools to analyze them provide unprecedented power for phylogenetic inference. Targeted sequencing has emerged as a strong tool for estimating species trees in the face of rapid radiations, lineage sorting, and introgression. Evolutionary relationships in Cyperaceae have been studied mostly using Sanger sequencing until recently. Despite ample taxon sampling, relationships in many genera remain poorly understood, hampered by diversification rates that outpace mutation rates in the loci used. The C4 Cyperus clade of the genus Cyperus has been particularly difficult to resolve. Previous studies based on a limited set of markers resolved relationships among Cyperus species using the C3 photosynthetic pathway, but not among C4 Cyperus clade taxa. We test the ability of two targeted sequencing kits to resolve relationships in the C4 Cyperus clade, the universal Angiosperms-353 kit and a Cyperaceae-specific kit. Sequences of the targeted loci were recovered from data generated with both kits and used to investigate overlap in data between kits and relative efficiency of the general and custom approaches. The power to resolve shallow-level relationships was tested using a summary species tree method and a concatenated maximum likelihood approach. High resolution and support are obtained using both approaches, but high levels of missing data disproportionately impact the latter. Targeted sequencing provides new insights into the evolution of morphology in the C4 Cyperus clade, demonstrating for example that the former segregate genus Alinula is polyphyletic despite its seeming morphological integrity. An unexpected result is that the Cyperus margaritaceus-Cyperus niveus complex comprises a clade separate from and sister to the core C4 Cyperus clade. Our results demonstrate that data generated with a family-specific kit do not necessarily have more power than those obtained with a universal kit, but that data generated with different targeted sequencing kits can often be merged for downstream analyses. Moreover, our study contributes to the growing consensus that targeted sequencing data are a powerful tool in resolving rapid radiations.
- Artículos de Revista