People in the Computational and Molecular Population Genetics (CMPG) lab use molecular techniques, theoretical developments, and computer simulations to reconstruct the demographic history of populations and species from genetic data, and to test between alternative evolutionary scenarios.
We explore the genomic diversity of voles and humans in order to discover which genes have recently responded to selection, for instance to adapt to new environments.
We are also interested in quantifying the effect of range expansions and colonisation processes on genetic diversity, since these demographic events can lead to molecular signatures resembling those of selection.
We also develop and maintain computer programs to study and simulate the genetic diversity of populations, infer demographic parameters under complex scenarios, and detect loci under selection from genome scan.
The CMPG lab is affiliated to the Swiss Institute of Bioinformatics.
Current Biology - Transition from background selection to associative overdominance (AOD) promotes diversity in regions of low recombination
Gilbert, Pouyet et al. combined mathematical modelling, individual-based simulations and human genomic data analysis to investigate how evolutionary forces determine patterns of genomic diversity. They find that selection against recessive deleterious mutations in regions of low recombination does not always lead to reduced diversity (compared to regions of intermediate recombination rate) as predicted by background selection (BGS) models. Contrastingly, selection against (partially) recessive mutations can lead to elevated levels of diversity at linked neutral sites and mimic patterns expected under balancing selection. Stable maintenance of distinct haplotypes in regions of low recombination does not necessarily come from any heterozygous advantage (i.e., overdominance for fitness) at any one of the genetic loci in this region. Rather, selection against recessive deleterious variants maintains complementary haplotypes if they all carry at least one harmful mutation at different positions This form of selection creates an increase of diversity at linked neutral sites, a process called associative overdominance (AOD). They derive the conditions for this AOD effect to occur and identify windows of the human genome where AOD seem to prevail. Negative selection in regions of low recombination may thus create more complex patterns of diversity than previously expected, emphasizing the necessity to modify the standard models for the effect of selection on linked neutral variation in low recombination regions to include harmful recessive mutations.