Rate and breadth of protein evolution are only weakly correlated

There is a weak positive correlation between rates and breadths of evolution, both across individual amino acid sites and across proteins.Rate and breadth are two distinct, and only weakly correlated, characteristics of protein evolution. The most likely explanation of their positive correlation is heterogeneity of selective constraint, such that less functionally important sites evolve faster and can accept more amino acids.This article was reviewed by Eugene V. Koonin, Arcady R. Mushegyan, and Eugene I. Shakhnovich.Evolution of a protein, at a particular site, can be characterized from two different perspectives [1,2]. The first is the rate of evolution, defined as the number of amino acid replacements which occur per unit of astronomical time or per time required for one selectively neutral substitution. The second is the "breadth" of evolution, i. e., the diversity of acceptable amino acids, which can be defined as the probability that two independent replacements of a particular amino acid at a site lead to the same amino acid.A priori, one could expect the rate and the breadth of protein evolution to be positively correlated across sites. Indeed, let us assume, in the spirit of the neutral theory [3], that, at a site, some amino acids are permitted, and confer the same fitness, and other amino acids are forbidden, because they confer very low fitnesses. Then, a site with a large number of permitted amino acids should display both a high rate and a wide breadth of evolution. Here, we investigate the rate-breadth correlation using the data on orthologous proteins from Drosophila and Mammalia.We used two datasets: 13 genomes of placental mammals (all the available genomes with coverage above ×5) with Monodelphis domestica as an outgroup, and 11 genomes of species from genus Drosophila with Anopheles gambiae as an outgroup. Genome-size multiple alignments of 44 vertebrates and 13 insects were downloaded from UCSC Genome Browser database [4], and the subsets of gen