For our next guest post Philipp Messer and Dmitri Petrov write about their paper
The McDonald-Kreitman Test and its Extensions under Frequent Adaptation: Problems and Solutions, arXived here
The McDonald-Kreitman (MK) test is the basis of most modern approaches to measure the rate of adaptation from population genomic data. This test was used to argue that in some organisms, such as Drosophila, the rate of adaptation is surprisingly high. However, the MK test, and in fact most of the current machinery of population genetics, relies on the assumption that adaptation is rare so that the effects of selective sweeps on linked variation can be neglected. We test this assumption using a powerful forward simulation and show that the MK test is severely biased even when the rate of adaptation is only moderate. The biases arise from the complex linkage effects between slightly deleterious and strongly advantageous mutations. In order to deal with these biases, we suggest a new robust approach based on a simple asymptotic extension of the MK test.
We further show that already under very moderate amounts of adaptation, linkage effects from recurrent selective sweeps can profoundly affect key population genetic parameters, such as the fixation probabilities of deleterious mutations and the frequency distributions of polymorphisms. In synonymous polymorphism data, these linkage effects leave signatures that can easily be mistaken for the signatures of recent, severe population expansion.
The bigger claim of our paper is that the effects of linked selection cannot be simply swept under the rug by introducing effective parameters, such as effective population size or effective strength of selection, and then using these effective parameters in formulae derived from the diffusion approximation under the assumption of free recombination. Given that most of our estimates of the key evolutionary parameters are still obtained from methods based on this paradigm, we argue that it is crucial to verify whether they are robust to linkage effects.
Philipp Messer and Dmitri Petrov