Vitor Pavinato has been recruited as postdoc for 18 months for the project ABCSelection (Detection of loci under selection from temporal population genomic data through ABC random forest) funded by the joint call from the LabEx AGRO, NUMEV and CEMEB. He will be supervised by Miguel Navascués and Jean-Michel Marin (Université de Montpellier).
One of the goals of population genetics is understanding the biological processes (migration, selection, etc.) that determine the genetic diversity of populations. To this end, model-based statistical inference is often used to characterize these forces from genetic data. Of particular interest are data of genetic diversity taken at several time points, which allow to track genetic changes through time and are more informative about the current processes acting on the population. With the advent of high throughput sequencing technologies the amount and, more importantly, the nature of the genetic data obtained have changed which requires new statistical approaches to analyze them. As part of his postdoc, Vitor will evaluate the use of random forests on the approximate Bayesian computation framework for tackling the analysis of population genomic data, in particular to co-estimate demography and selection from time-series population genomic data.