Boussau Bastien

My past research has been focused on extracting information from genomes to better understand how they encode phenotypes. Extant living organisms are the result of an historical process that has unfolded over billions of years. Their genomes have accumulated footprints of past episodes of selection in response to interactions with their environment or with other species. I have developed computational methods based on probabilistic models or on deep neural networks to detect these footprints and interpret genomic data on a large scale.

In particular, I have used ancestral sequence reconstruction to study the lifestyles of organisms that lived billions of years ago, I have developed methods for reconstructing phylogenies, for dating phylogenies, for jointly reconstructing gene trees and species trees, and I have developed and assessed methods to find examples of convergent genomic evolution.

My research in genomics these days mainly focuses on:

•  using deep learning approaches to help in the fields of phylogenetic reconstruction and molecular evolution (collaborations with Enzo Marsot, Damien de Vienne, Philippe Veber, Luc Blassel and Laurent Jacob)
• detecting directional selection at the sequence level, notably to study convergent genomic evolution (collaborations with Estelle Bergiron, Pierre Gérenton, Vincent Lacroix, and Philippe Veber)
• exploiting information from horizontal gene transfers to date species phylogenies (e.g. see this manuscript and the associated recommendation by PCI Evol Biol)

I have advised several students in the past, including Carine Rey, Adrián A. Davín, Baptiste Maucourt, Louis Duchemin, Johanna Trost, Luca Nesterenko.

In terms of teaching, I have been involved in teaching Bayesian statistics, Computational Molecular Evolution, Phylogenetic Reconstruction, Genomics.

I also teach about the environmental footprint of our food systems as part of the course "Climat et transitions" at Université Lyon 1.