4: Why are some phylogenetic relationships easy and others hard to resolve? A phylogenomic case study in ants

Many phylogenomic studies find that while some relationships are easily resolved and have held up since Sanger sequencing days, many difficult nodes remain so even with genome-scale data. Despite theory predicting that incomplete lineage sorting, introgression, and systematic bias are likely to produce discordant or unreliable phylogenetic estimates within genomes and across species, early phylogenomic work was conducted with the hope that more data will help resolve even the most recalcitrant nodes. With a number of whole-genome studies uncovering discordant relationships now published, the focus is shifting to investigatigation of sources of discordance within genomes and across the tree of life. Here we analyze a new phylogenomic dataset for ants, sequencing ultraconserved elements in 177 genera and combining those with publicly available sequences for a total of 271 taxa together corresponding to 82% of described genera and representing 98% of described species-level diversity. This phylogeny presents examples of both “easy” and “difficult” nodes. We investigate the relative contribution of biological and technical causes behind poorly resolved relationships using simulations and recently developed approaches to measuring phylogenomic support.