There has been some discussion on the effect of population size on positive selection scans. It has been argued that positive selection as well as purifying selection is less efficient in humans and chimps than rodents because of their smaller population sizes than murids (Keightley et al., 2005; Rhesus Macaque Sequence and Analysis Consortium, 2007).
We have performed a positive selection scan on six mammalian genomes: human, chimpanzee, mouse, rat and dog. We find fewer positive selected genes on the primate clade than on the rodent clade. We try to quantify this effect by estimating ratios of population sizes of the species from the genomic data. The estimation is based on a extension of population genetic interpretation of a branch model of codon substitutions by Halpern and Bruno 1998 (see also Nielsen and Yang, 2003). The estimates we obtain for the ratio of macaque and human effective population sizes are in good agreement with the ratio calculated from polymorphim data (Wall et al., 2003, Hernandez et al, 2007). Calculations of further population size ratios on the mammalian tree with our method are underway. We will also discuss limitations of the estimation of population ratios from genomes of single individuals.
view more