Please refer to https://doi.org/10.1534/genetics.117.300467 or contact Julien for further information on the projects: email@example.com
1. The evolution of stochastic gene expression
How a genotype is realized into a phenotype is the result of the process of gene expression. While long thought as a deterministic genetic "program", the advent of single cell biology firmly established gene expression as a stochastic process, which results from the diffusion and binding of relatively small numbers of molecules within the cell. High throughput single cell genomics technologies provided evidence that expression noise is under specific selective pressure, which depends on the function of the gene, as well as complex factors such as the position of the gene in the interaction network . In this project, we aim to study empirically how gene expression noise evolves. We will use natural single cell organisms, yeasts, and compare several species with different degrees of divergence. Using single-cell RNA sequencing, we will quantify gene- and species- specific expression noise. We will then develop dedicated phylogenetic models to study the patterns of expression noise evolution.
2. New methods for inferring the rate of recombination along genomes
Recombination is a fundamental evolutionary process: as it restores linkage equilibrium, it impacts genetic diversity and modulates the efficacy of selection. Recombination rate was shown to vary greatly between and within species, and recombination maps are essential tools for geneticists, as they are key ingredients of association studies, as well as selection scans. The "standard" methods to infer recombination maps from population genomic data consist in predicting the recombination rate between any two SNPs, based on their pattern of linkage disequilibrium. Such methods typically require a large sample size to properly infer the recombination rate. Recently , we developed a new modeling framework based on the sequentially Markov coalescent (SMC) that enables the reconstruction of recombination maps from a pair of genomes only. In this project, we propose to extend this method to efficiently combine the information of multiple genomes and gain increased resolution. We will also apply the method to the reconstruction of ancestral recombination maps in order to study the evolution of the recombination process.