Fig. 6

Phylogenetic distribution (PD) of gene families partially explains gene family variability. Scatter plot shows log10 PD (x-axis) vs. log10 residual variance statistic (y-axis). Red points were significantly variable and blue points were significantly invariable. Gene families in specific functional groups are also highlighted in different colors, specifically the bacterial ribosome (green), the type VI secretion system (or “T6SS”; orange), the KinABCDE-Spo0FA sporulation control two-component signaling system (yellow), and hypothetical genes (tan squares). Gene families that were significantly invariable (ribosome and sporulation control) or significantly variable (hypothetical genes and the T6SS) at an estimated 5% FDR are outlined in black. The bacterial ribosome, as expected, had very high PD and was strongly invariable. The type VI secretion system genes, in contrast, were conserved but variable, and some genes involved in the Kin-Spo sporulation control two-component signaling pathway had low PD but were invariable. Only gene families with at least one annotated bacterial or archaeal homolog are shown