 |
Center for Nonlinear Studies |
When infectious diseases spread through heterogeneous host populations, non-random contact patterns can produce measurable changes in genetic diversity of the pathogen. I consider two related problems: Given pathogen sequence data sampled from different risk groups or spatial locations in a host population, how may we estimate transmission rates between risk groups? And, if we do not observe the risk group or location of sampling of pathogen sequences, is it possible to detect the presence of host population structure and different risk groups? I describe a new approach to detecting latent structure in pathogen phylogenies and show how this can be used to detect outbreaks and transmission hotspots. These methods are applied to Ebola virus in Western Africa, Human Immunodeficiency Virus (HIV) in Tennessee, and Neisseria gonorrhoeae in North America. Each of these examples demonstrates the potential to detect known population structure and indicates the presence of additional structure which has not yet been characterized or explained.
Host: Thomas Leitner