Section: New Results

Genome histories reconstructions

E. Tannier (Beagle), with B. Boussau, G. Szollosi, V. Daubin, L. Duret, M. Gouy, S. Abby (LBBE, Lyon), N. Lartillot (Univ Montreal), C. Chauve (SFU Vancouver)

Lateral gene transfer has been discovered in the 1940's and since has been seen by phylogeneticists as a noise one had to remove before analyses in molecular evolution. Thi noise was recently considered so important that it would blur the historical signal and leave no hope for reconstructing a phylogeny. In a series of papers [16] , [31] , [32] , we model the lateral gene transfer and prove that it can be used as a signal to

• reinforce the support for the phylogeny of vertical descent [16]

• order in time some bacterial diversification events, and thus provide a unique source for dating the history of life (more than 3/4 of it is prokaryotic and the fossil record is not abundant) [31]

• have a trace of extinct species which did not leave any descendants, if they gave some genes to more successful lineages, which opens the way to include them in molecular phylogenies [32]

We devised methods to trace whole genome evolution, with multi-scale mutations: from nucleotide substitutions to large-scale rearrangements. We provided a mammalian phylogeny accounting fot the evolution of several thousand genes [17] , and a method to sample among evolutionary scenarios [27] .

Eventually we built a model of evolution of relations between pairs of genes, enable us to reconstruct ancestral genome structure or ancestral systems of interactions [18] . In the case of genome structure we also published a method to linearize a set of ancestral relations [26] .