Breen et al. reply

replying to D. M. McCandlish, E. Rajon, P. Shah, Y. Ding & J. B. Plotkin Nature 497, http://dx.doi.org/10.1038/nature12219 (2013) Understanding fitness landscapes, a conceptual depiction of the genotype-to-phenotype relationship, is crucial to many areas of biology. Two aspects of fitness landscapes are the focus of contemporary studies of molecular evolution. First, the local shape of the fitness landscape defined by the contribution of individual alleles to fitness that is independent of all genetic interactions. Second, the global, multidimensional fitness landscape1 shape determined by how interactions between alleles at different loci change each other’s fitness impact, or epistasis. In explaining the high amino-acid usage (u), we focused on the global shape of the fitness landscape2, ignoring the perturbations at individual sites3.