pleiotropy - senecio7の日記

Lourenço J, Galtier N & Glémin S 2011 Complexity, pleiotropy, and the fitness effect of mutations. Evolution 65:1559-1571.

even in a long-term stable environment, populations are never perfectly adapted
because of the stochastic fixation of slightly deleterious mutations, phenotypes are expected to be distributed around the optimum, at a mean phenotypic distance that results from a dynamic equilibrium between mutation, selection, and drift
the resulting decrease in mean population fitness is commonly known as the drift load, and is inversely related to the effective population size
drift load is directly proportional to complexity
because fitness is a measure of how successfully organisms interact with the environment, a higher number of these interactions can make simultaneous optimization more difficult

we can calculate the mean fitness of the population at equilibrium
w_eq = exp(− n / (4N_e)) ... (14)
drift load, while proportional to complexity and inversely proportional to effective population size, is not expected to depend on m (or σ)
when the complexity of organisms increases, the proportion of mutations that are deleterious increases also, and so the mean distance to the optimum tends to increase until the balance between the effects of fixed deleterious and advantageous mutations is reestablished
when effective population size increases, there are fewer effectively neutral deleterious mutations (the ones with a fair chance of being fixed), and so the distance to the optimum at equilibrium tends to decrease until the balance is reestablished
when populations are so small that drift overwhelms selection, there is a higher frequency of effectively neutral deleterious mutations that fix, causing fitness to decline, until the increase in the rate of fixation of advantageous mutations reestablishes the equilibrium
an effect termed compensatory epistasis (Silander et al. 2007)

very large populations maintain nearly optimal mean population fitness regardless of phenotypic complexity
small populations can maintain high mean fitness only when there are a small number of traits

the fact that drift load is largely independent of the DFEM in FGM, is valid under the assumption of alleles having continuous effects
under alternative models (such as a model where alleles have discrete effects), the drift load may depend on the fitness effect of mutations

drift load depends only on complexity (n)
the DFEM at mutation-selection-drift balance depends only on mutational pleiotropy (m)

Wagner et al. (2008b) present evidence of an increase in the total effect of QTL with increasing degree of pleiotropy, that is even stronger than predicted by the ESM model
this evidence lead them to reject the ESM model
as pointed out in Hermisson and McGregor (2008), the authors did not consider the possibility of some of the analyzed QTL regions containing multiple mutations, which could lead to an overestimation of the effects of mutations with higher pleiotropy
in a reply, the authors acknowledged that these overestimations can reconcile the observed data with the ESM model
so this model remains a reasonable assumption

the distinction between mutation pleiotropy and total pleiotropy (or complexity), could explain much of the discrepancy between the studies done by Martin and Lenormand (2006) and Tenaillon et al. (2007)
the former approach essentially estimating m, and the latter n