PHYPD · New phylogenetic methods for inferring complex population dynamics

The field of phylogenetics is undergoing an important change over recent time. For a long time phylogenetics focussed on describing the species tree as accurate as possible. More recently there is a shift towards a functional inference of the evolutionary process itself, i.e. researchers attempt to infer what processes are most likely responsible for the particular pattern of branching observed in a given phylogeny.Phylogenetics has traditionally been developed for macroevolution. Recently, it has been recognized that the phylogenetic concept is also crucial for improving our understanding of epidemiological processes. In an epidemiological phylogeny, tips of the tree are corresponding to infected hosts (instead of extant species), and branching events correspond to transmission events (instead of speciation events).In the first part of the project, I propose to develop novel phylogenetic methodology that will allow us to characterize fundamental evolutionary processes based on species phylogenies. I will aim at identifying key factors (such as environment, competition between species, or microevolutionary processes) determining macroevolutionary dynamics. Insight into these dynamics will allow us to address fundamental evolutionary questions such as the advantages of recombination as well as to revisit current debates about the impact of climate change on diversity.In the second part of the proposed project, I will develop tools identifying the main drivers responsible for the spread of an epidemic, which may help informing public health intervention strategies. Furthermore, I aim at characterizing bottlenecks at transmission, which may become important for early treatment strategies or vaccine design.Completion of the projects will deepen our understanding of macroevolutionary and epidemiological dynamics, as well as lead to novel phylogenetic tools allowing us to analyse the growing amount of available data, such as next-generation sequence data.