Premise : Discordance between nuclear and organellar phylogenies (cytonuclear discordance) is a well-documented phenomenon at shallow evolutionary levels, but has been poorly investigated at deep levels of plant phylogeny. Determining the extent of cytonuclear discordance across major plant lineages is essential not only for elucidating evolutionary processes but also for evaluating the currently used framework of plant phylogeny, which is largely based on the plastid genome.

Methods : We present a phylogenomic examination of a major angiosperm clade ( Asteridae ) based on sequence data from the nuclear, plastid, and mitochondrial genomes as a means of evaluating currently accepted relationships inferred from the plastome and exploring potential sources of genomic conflict in this group.

Results : We recovered at least five instances of well-supported cytonuclear discordance concerning the placements of major asterid lineages (i.e., Ericales, Oncothecaceae, Aquifoliales, Cassinopsis , and Icacinaceae). We attribute this conflict to a combination of incomplete lineage sorting and hybridization, the latter supported in part by previously inferred whole-genome duplications (WGDs).

Conclusions : Our results challenge several long-standing hypotheses of asterid relationships and have implications for morphological character evolution and for the importance of ancient WGDs in early asterid evolution. These findings also highlight the value of reevaluating broad-scale angiosperm and green plant phylogeny with nuclear genomic data. 

This dataset comprises nuclear, plastid, and mitochondrial gene sequences generated through a variety methods (target enrichment, genomic skimming, transcriptome sequencing, and sanger sequencing). Many of the sequences were obtained from publically available sources (1KP, GenBank). New sequences were generated using target enrichment (for nuclear genes) and genome skimming on the Illumina sequencing platform. Methods of gene assembly, alignment, and phylogenetic analysis are outlined in the paper.

Three alignments are provided, one each for the nuclear, chloroplast, and mitochondrial genome. Each (concatenated) alignment file has an assocaited partition file listing the starting and stopping positions of each gene. This can be used to split the concatenated alignements into individual gene alignments as needed.