Evolution of genes neighborhood within reconciled phylogenies: an ensemble approach
Résumé
CONTEXT:The reconstruction of evolutionary scenarios for whole genomesin terms of genome rearrangements is a fundamental problem in evolutionaryand comparative genomics. The DeCo algorithm, recently introducedby Berard et al., computes parsimonious evolutionary scenarios forgene adjacencies, from pairs of reconciled gene trees. However, asfor many combinatorial optimization algorithms, there can exist manyco-optimal, or slightly sub-optimal, evolutionary scenarios thatdeserve to be considered.CONTRIBUTION:We extend the DeCo algorithmto sample evolutionary scenarios from the whole solution space underthe Boltzmann distribution, and also to compute Boltzmann probabilitiesfor specific ancestral adjacencies.RESULTS:We apply our algorithmsto a dataset of mammalian gene trees and adjacencies, and observea significant reduction of the number of syntenic conflicts observedin the resulting ancestral gene adjacencies.
Domaines
Bio-informatique [q-bio.QM]Origine | Publication financée par une institution |
---|