Populations of RNA viruses are often characterized by abundant genetic variation. However, the relative fitness of these mutations is largely unknown, although this information is central to our understanding of viral emergence, immune evasion, and drug resistance. Here we develop a phylogenetic method, based on the distribution of nonsynonymous and synonymous changes, to assess the relative fitness of polymorphisms in the structural genes of 143 RNA viruses. This reveals that a substantial proportion of the amino acid variation observed in natural populations of RNA viruses comprises transient deleterious mutations that are later purged by purifying selection, potentially limiting virus adaptability. We also demonstrate, for the first time, the existence of a relationship between amino acid variability and the phylogenetic distribution of polymorphisms. From this relationship, we propose an empirical threshold for the maximum viable deleterious mutation load in RNA viruses.

Original publication

DOI

10.1093/molbev/msm001

Type

Journal article

Journal

Mol Biol Evol

Publication Date

03/2007

Volume

24

Pages

845 - 852

Keywords

Amino Acid Sequence, Computational Biology, Evolution, Molecular, Likelihood Functions, Models, Genetic, Mutation, Phylogeny, Polymorphism, Genetic, RNA Viruses, Selection, Genetic