RESEARCH

The factors that determine selection efficacy are poorly understood. A proposed factor is the interaction between mutations’ dominance and ploidy level: recessivity should reduce selection efficacy in diploids but not in haploids. Methodological limits make this challenging to test in species with a diploid-dominant life cycle. Yet, a whole phylogenetic clade under-explored in genomics gathers all the attributes to tackle this question: Bryophytes. Their life cycle is characterized by alternating a long haploid phase (gametophyte) with a short diploid phase (sporophyte), both phases being macroscopic.

This is ideal for understanding how ploidy affects the efficacy of adaptive selection, purifying selection and selection against hybridization. And to shed light on the evolution of haploid sex chromosomes.

Speciation is a process of gradual accumulation of reproductive barriers in genomes, ultimately leading to a cessation of gene flow between groups of individuals forming distinct biological species. This barrier effect to interspecific gene flow is a major concept for the study of speciation, as it makes it possible to quantify the intensity of reproductive isolation using population genomics. Another major advance in the genetics of speciation has been the study of the role of sex chromosomes. But these two lines of research have been conducted largely independently. 

Our goal is thus to quantify the barrier effect played by sex chromosomes of XY and ZW sexual systems in relation to autosomes. This is done by comparing their rate of interspecific introgression through modelling and genomic data analysis.

Genome scans of differentiation in a variety of taxa have shown that levels are heterogeneous across the genome of incipient species. This observation is a source of debate as it could be due to differences in effective migration rates between "barriers" and "non-barriers" regions, or differences in the strength of linked selection. Thus it is crucial to go beyond genome scans to understand the impact of gene flow on the evolutionary trajectories of incipient species.

We develop flexible and scalable statistical methods based on machine learning techniques to distinguish between various scenarios of speciation while taking into account the effect of linked selection on genomic variation.

Dobzhansky and Muller developed a model of epistasis between pairs of loci (DMIs) that easily explains the evolution of reproductive isolation, without the need for populations to cross fitness valleys.  However, speciation is a complex, and probably multigenic, process that may require the interactions between a large number of loci.

Thus, it is essential to develop more realistic models that account for the effects of multiple barrier loci. We develop such models of divergence between allopatric populations using a variety of theoretical frameworks (local adaptation models, ad hoc epistasis models, Fisher’s geometric models) to understand how incompatibilities accumulate between species, and what are their long-term effect as barriers to neutral gene flow during secondary contacts.