Coalescent-based models for selection and superspreading in viral phylogenies Patrick Hoscheit October 12th, 2015 Patrick Hoscheit SelPhy October 12, 2015 1 / 6
Phylodynamics Principle The epidemiology of a virus is strongly linked to its evolution. We can use one to understand the other. Starting from a sample of viral genetic sequences, Estimate a phylogeny Write down a model relating the epidemic to the observed phylogeny (possibly including other data) Estimate model parameters through Bayesian or Maximum-Likelihood methods Usually, all three steps are performed together to take uncertainty in phylogenetic reconstruction into account (BEAST). Patrick Hoscheit SelPhy October 12, 2015 2 / 6
Example : Gire et al. (2014) Patrick Hoscheit SelPhy October 12, 2015 3 / 6
Superspreading and selection Superspreading refers to non-heritable exceptional transmission events. Distorts phylogenies by creating many closely related lineages Many pathogens are concerned (Ebola, MERS-CoV, FMD,…) Strongly related to “environmental” conditions (contacts, breeding, sanitary measures,…) Selection acts on pathogens mostly through cross-immunity, but also through sanitary interventions (vaccination, antibiotics,…) Lineages with high fitness tend to replace other ones Creates highly imbalanced phylogenies Patrick Hoscheit SelPhy October 12, 2015 4 / 6
Example : Inter-patient HIV vs. Influenza A Patrick Hoscheit SelPhy October 12, 2015 5 / 6
Our project We aim to find new methods to deal with superspreading and strong selection in viral phylogenies by using nonbinary coalescent models For now, only applicable to (RNA) viruses, but maybe in the future, bacteria or eukaryotes In collaboration with Prof. Oliver Pybus (Dept of Zoology, Univ. of Oxford) March 2015–March 2016 Come see my poster ! Patrick Hoscheit SelPhy October 12, 2015 6 / 6
Recommend
More recommend