Download seminar poster
Talk Title :
Phylogenomics of animals

Date / Time / Location:
Thursday March 12^th, 2009 - 6:00 pm
Room 232, Leacock Building
McGill University
855 Sherbrooke Street West

Affiliation :
Centre Robert-Cedergren

URL
Hervé Philippe

Abstract :
The metazoan phylogeny has undergone major rearrangements at the end of the nineties. However, the resolution was rather poor, preventing to draw any firm conclusions. Phylogenomics, with the availability of large sets of sequences, is usually expected to increase the resolving power of molecular phylogenetic inference, leading potentially to fully resolved tree. Unfortunately, numerous phylogenomic trees of animals have been recently published and not only resolution is often poor and but also results are regularly contradictory. I will explore the reasons of these apparently deceiving results:

1. Inadequate selection of sequences (non-orthologous sequences, saturated genes).
2. Inadequate taxon sampling.
3. Inadequate tree reconstruction method. When more and more data are considered, although stochastic error vanishes, systematic error becomes more apparent. Such errors are due to the imperfection of models of sequence evolution.
4. Rapid diversification of animal phyla (Cambrian explosion). Lack of resolution is generally interpreted as evidence of closely spaced speciation events. Molecular dating may allow to test this hypothesis.

In conclusion, I will present my view on the phylogenetic relationships of animals.

Download seminar poster
Talk Title :
Vertebrate Comparative and Regulatory Genomics

Date / Time / Location:
Thursday February 12^th, 2009 - 6:00 pm
Atrium, Bellini Building
McGill Life Sciences Complex

Affiliation :
MCB, McGill Center for Bioinformatics

URL
Mathieu Blanchette

Abstract :
This talk will bring together my two passions, genome evolution and gene regulation, and will discuss what the former can teach us about the latter. We have recently shown that the genome of an ancestral mammal living about 70 Million years ago can be computationally reconstructed to a surprising degree of accuracy from the genomes of extant species. In the first part of this talk, I will introduce some of the computational challenges related to the accurate reconstruction of an ancestral genome. I will then describe how this new genome, strategically positioned at the center of the mammalian radiation, allows a detailed study of the evolution of transcriptional regulatory regions. Finally, I will describe how whole-genome analyses based in part on comparative genomics reveal new and unexpected traits of mammalian regulatory regions, both those located in non-coding regions and those hiding within protein-coding DNA.