Lawrence Berkeley National Laboratory
A Theoretical Lower Bound for Selection on the Expression Levels of Proteins.
- Author(s): Price, Morgan N
- Arkin, Adam P
- et al.
Published Web Locationhttps://doi.org/10.1093/gbe/evw126
We use simple models of the costs and benefits of microbial gene expression to show that changing a protein's expression away from its optimum by 2-fold should reduce fitness by at least [Formula: see text], where P is the fraction the cell's protein that the gene accounts for. As microbial genes are usually expressed at above 5 parts per million, and effective population sizes are likely to be above 10(6), this implies that 2-fold changes to gene expression levels are under strong selection, as [Formula: see text], where Ne is the effective population size and s is the selection coefficient. Thus, most gene duplications should be selected against. On the other hand, we predict that for most genes, small changes in the expression will be effectively neutral.