| dc.description.abstract | The close coexistence between human beings and dogs and the efficiency in their domestication have made dogs a very interesting target for basic research. By studying the history and evolution of dogs, much can be learned about human health, history and social behavior. Moreover, dogs show some astounding diversity patterns which make them mystifyingly unique; to date, no other species has ever been found to spawn such a huge morphological variability (dog breeds) from a very little initial effective population size in a very short period of time.
In this project, we have put all our efforts into two noteworthy aspects. First, we completely redesigned an in-house mitochondrial DNA reconstruction pipeline which, after having assessed its efficiency and performance, we used it to reconstruct a large dataset containing almost all the available dog breeds and well as other samples from different canids. Then, the reconstructed sequences were used to study crucial aspects of dog evolution: A massive scale phylogeny was built to see whether the current morphology-based classification of dogs truly correlates with their genetic distances as well as other aspects. In addition to that, we have inferred the population history of dogs, which showed a population bottleneck and posterior expansion attributable to a domestication event, and we got results about the Time to the Most Common Ancestors (TMRCAs) between divergent species. Furthermore, we assessed some evolutionary statistics which showed the strong purifying selection acting on canids, but also that dogs could have a relaxation in this selection due to the artificial breeding and selection made by humans.
Finally yet importantly, we encountered a discrepancy in the starting codon of the ND4L gene in the reference genome of dogs (CanFam 3.1.) which suggest that the gene does not start in the proposed codon but in a nearby one.
Key words: bioinformatics, mtArchitect, dogs, canids, wolves, phylogenesis, mitochondrial DNA. | es |
