Abstract:
|
The goal of phylodynamics is to estimate effective population size trajectories from genetic sequence data. Modern nonparametric Bayesian methods offer the state-of-the-art for recovering population size trajectories of unknown form, but current methods cannot accurately recover trajectories that exhibit abrupt changes or varying levels of smoothness. We propose a novel, locally-adaptive approach to Bayesian nonparametric phylodynamic inference which has the flexibility to accommodate a large class of functional behaviors in the underlying population size trajectory. Recent advances in phylodynamic inference account for differential sampling intensities as functions of population size. We build on these models by allowing both the population size trajectory and the functional relationship between sampling intensity and population size to vary with time via our locally-adaptive smoothing method. We use simulated data to assess model performance, and find that our proposed method results in reduced bias and increased precision when compared to contemporary methods. We also apply our models to real data from seasonal human influenza outbreaks.
|