In the second half of the twentieth century, gene sequences have been used to study their evolution as well as the history of species. A decade ago, progress in sequencing techniques provided the opportunity to use whole genome sequences for reconstructing species phylogeny. Several phylogenomic methods have been used to exploit these large amounts of data, and those that are based on homologous or orthologous characters have been favoured. Among these, the supertree and supermatrix approaches have been very successful at resolving parts of the tree of life. However, these methods discard a great proportion of the evolutionary information present in genomes. New integrative models of genome evolution could make a better use of whole genome sequences and thus improve the resolution of the tree of life.