Woolfit, M., & Bromham, L. (2003). Increased rates of sequence evolution in endosymbiotic bacteria and fungi with small effective population sizes. Molecular Biology and Evolution, 20(9), 1545-1555.
Woolfit and Bromham (2003), "Increased rates of sequence evolution in endosymbiotic bacteria and fungi with small effctive population sizes", discusses the previously found evidence supporting the idea that endosymbiotic bacteria experience lower effective population sizes than their free-living relatives, and then demonstrates that phylogenetic independence of samples in these previous studies was not considered, that the 16S gene is not a good demonstration of base bias compared to a whole genome sequence, and that phenomenon such as relaxed selection and increased mutation rate appear to be inadequate compared to the hypothesis of reduced population size. The authors use a variety of computational and statistical methods to establish their findings, including phylogenetic analysis paired with maximum likelihood and LogDet, a variety of methods for determining the relative rates of substitution, A+T content bias, and a lot of signed-rank analyses. The authors used available endosymbiont gene/genome data to demonstrate that phylogenetic independence is necessary for accurate statistical outcomes and that the patterns observed in the endosymbiont genomics are likely due to reduced population size as opposed to other explanations.
Can the patterns being revealed be used to identify unique or interesting endosymbiont genome hidden within databases, perhaps via machine learning?
What does the lack of cell-surface component biosynthesis genes (in the endosymbiont) mean for the discussion of host-microbe interactions and the determination of self vs. non-self? Is the bacterium considered self now by the host? What key stages might lead to this point, or further?
Are hosts preferentially providing A+T to endosymbionts due to lower costs for producing those nucleotides? Could we do an experiment wherein C+G is already limited, and then see where added C+G end up-- perhaps using radiolabelling.