Data from: Comparison of target-capture and restriction-site associated DNA sequencing for phylogenomics: a test in cardinalid tanagers (Aves, genus: Piranga)

Restriction-site associated DNA sequencing (RAD-seq) and target capture of
specific genomic regions, such as ultraconserved elements (UCEs), are
emerging as two of the most popular methods for phylogenomics using
reduced-representation genomic datasets. These two methods were designed
to target different evolutionary timescales: RAD-seq was designed for
population-genomic level questions and UCEs for deeper phylogenetics. The
utility of both datasets to infer phylogenies across a variety of
taxonomic levels has not been adequately compared within the same
taxonomic system. Additionally, the effects of uninformative gene trees on
species tree analyses (for target capture data) have not been explored.
Here, we utilize RAD-seq and UCE data to infer a phylogeny of the bird
genus Piranga. The group has a range of divergence dates (0.5 my – 6 my),
contains eleven recognized species, and lacks a resolved phylogeny. We
compared two species tree methods for the RAD-seq data and six species
tree methods for the UCE data. Additionally, in the UCE data, we analyzed
a complete matrix as well as datasets with only highly informative loci. A
complete matrix of 189 UCE loci with ten or more parsimony informative
(PI) sites, and an ~80% complete matrix of 1128 PI SNPs (from RAD-seq)
yield the same fully resolved phylogeny of Piranga. We inferred
non-monophyletic relationships of P. lutea individuals, with all other a
priori species identified as monophyletic. Finally, we found that species
tree analyses that included predominantly uninformative gene trees
provided strong support for different topologies, with consistent
phylogenetic results when limiting species tree analyses to highly
informative loci or only using less informative loci with concatenation or
methods meant for SNPs alone.