Regressions showing red algal signal in diatoms is not shared with oomycetes. A. Regressions on second hits for genes that are present in all eukaryotic groups (therefore, unrelated to plastid function), where the top hit is to the sister heterokont group (e.g. hit to oomycetes when diatoms are query sequences). The query genome in each case is shown in the upper right corner of the plot. Broken lines represent quadratic and solid lines linear regressions with adjacent R2 values shown. In genes most similar between the heterokont sister groups, there is no apparent phylogenetic signal from red algae in either oomycete or diatom genomes; that is, hits to reds do not deviate positively from the value predicted by the regression model. B. Conversely, with oomycetes removed from the analysis, a regression on top hits versus group size clearly shows a positive signal for red algal genes. This same pattern was found in regressions on top hits against group size for all groups present (Additional file 2, Table 5). Contrary to predictions of the chromalveolate hypothesis, these comparative analyses indicate that the clearly detectable red algal signal in diatom genomes is not present in genes specifically shared with oomycetes.
Stiller et al. BMC Genomics 2009 10:484 doi:10.1186/1471-2164-10-484