Understanding the evolution of complex biological processes involving many genes and proteins poses problems similar to those encountered in statistical physics, where methods have been developed to describe phenomena emerging from a large number of interacting entities. I will discuss how ideas from statistical physics can be applied to the evolutionary dynamics of such ‘many gene problems’. Whereas the evolution of plants and animals took hundreds of millions of years, viruses can change significantly over a few years, the ‘swine flu’ outbreak this spring providing a recent example. While the rapid evolution of viruses is a threat to human health, it provides us with the unique opportunity to study the general features of evolution, either through planned experiments or analyzing how the viral genomes change over the years. Of particular interest is the role of recombination, a form of viral sex in which different viruses mix their genes. Recombination produces new individuals from existing genetic variation, which is believed to speed up evolution and in the case of viruses can give rise to more virulent strains of flu or drug resistant strains of HIV. However, the quantitative effects of recombination on evolution are rather poorly understood, in particular when many genes are involved. I will discuss how sex speeds up evolution and how the interplay between reshuffling of genetic variation during sexual reproduction and interaction between genes can give rise to qualitatively different regimes of evolutionary dynamics.
