Population genetics and substitution models of adaptive evolution

TL;DR

This paper critically examines the use of the ratio of non-synonymous to synonymous substitutions ($$) as a measure of adaptive evolution, proposing new models and scenarios where $$ can exceed 1, especially after environmental changes.

Contribution

It introduces models of amino acid substitution that account for changing selection pressures and provides a framework for interpreting $$ in various evolutionary contexts.

Findings

$<1$ under constant selection.

Transient $ \u226a 1$ after environmental shifts.

Expected $>1$ at equilibrium under frequency-dependent selection.

Abstract

The ratio of non-synonymous to synonymous substitutions $\omega(=d_{N}/d_{S})$ has been widely used as a measure of adaptive evolution in protein coding genes. Omega can be defined in terms of population genetics parameters as the fixation ratio of selected vs. neutral mutants. Here it is argued that approaches based on the infinite sites model are not appropriate to define $\omega$ for single codon locations. Simple models of amino acid substitution with reversible mutation and selection are analysed, and used to define $\omega$ under several evolutionary scenarios. In most practical cases $\omega<1$ when selection is constant throughout time. However, it is shown that when the pattern of selection on amino acids changes, for example after an environment shift, a temporary burst of adaptive evolution ($\omega\gg1$) can be observed. The fixation probability of a novel mutant under frequency dependent selection is calculated, and it is used to show why $\omega>1$ can be sometimes expected for single locations at equilibrium. An example with influenza data is discussed.