Experimental design and statistical rigor in phylogenomics of horizontal and endosymbiotic gene transfer

Although specific details of how plastids originated and spread among eukaryotes remain under debate [1-8], there is little doubt that extant photosynthetic taxa evolved through a very complicated process. There have been several independent primary origins of plastids from cyanobacterial endosymbionts [9,10], as well as undetermined numbers of secondary, tertiary, and perhaps higher order endosymbioses involving eukaryotic to eukaryotic plastid transfer (see [11-13] for reviews). If hypotheses that minimize the number of endosymbioses prove correct, plastids also have been lost on numerous occasions [1,14,15]. Finally, there have been a number of proposed and documented cases of serial replacement of endosymbionts from very different taxonomic sources [16-21].Regardless of which of these scenarios turn out to be validated, it is clear that plastid endosymbioses have remodeled host cell genomes substantially through the process of "endosymbiotic gene transfer (EGT). EGT is a special case of horizontal gene transfer (HGT) involving concerted movement of many endosymbiont genes into the host cell nucleus [22]; its impacts have been dramatic (Figure 1A). For example, nearly 20% of genes in the Arabidopsis nuclear genome are derived from the cyanobacterial ancestor of chloroplasts [23]. Moreover, only about half of these genes appear to be related to chloroplast function; the rest either replaced original eukaryotic homologs or were adapted to entirely novel functions in the host cell's metabolism. Overall contributions from EGT vary among photosynthetic taxa [24] and eukaryotic to eukaryotic EGT is more difficult to quantify [22], partly because phylogenetic relationships among major groups remain unclear. Nevertheless, there is no doubt that large scale EGT has been a feature of both primary and higher order plastid endosymbioses [22]. Transferred genes that are not related to plastid function are of particular interest because many should remain under strong purifyin