Collaborative Research: eMB: Next-generation phylodynamics: theory, algorithms, applications

About This Grant

Viruses experience frequent small random changes in their genetic material, their genomes. Because many of the changes in the viruses that infect one animal happen independently of those that happen in another animal, one can compare the genomes of sampled viruses to glean information about how far and fast an epidemic is spreading. This is known as phylodynamics. This project will develop new mathematical and computational tools to allow us to extract more information about how a virus is moving through a population of animals from virus genomes. Specifically, recent mathematical breakthroughs allow us to understand more precisely how aspects like virus transmission, severity of disease, and duration of immunity—and differences among animals in these aspects—leave their marks on virus genomes. The project will capitalize on these developments, along with recent advances in machine learning technology and the world’s premiere database of avian influenza virus genomes, to reduce some of the key uncertainties about how this virus spills over from wild birds into domestic animals, and potentially into humans. The project is expected to benefit public health by helping us better understand how avian influenza spreads and where the greatest risk-points are by increasing the usefulness of a very common kind of data. The mathematical and computational tools developed will also be useful in other scientific and medical fields, including cancer biology and microbiology. The project will develop a short-course for training epidemiologists and mathematical biologists in phylodynamic methods. Phylodynamics seeks to extract information from genomes of individuals to shed light on population-scale dynamic processes. Its development has largely been driven by applications in epidemiology, where pathogen genomes contain information concerning determinants of disease transmission. In this context, phylodynamics has become essential in guiding public-health response in epidemics at a variety of geographical and temporal scales. From the mathematical point of view, the aim of phylodynamics is to infer the structure and parameterization of mathematical models of demographic processes on the basis of accumulated differences among sampled genome sequences. Existing approaches rest on assumptions (large population sizes, small sample fractions, linearity of demographic processes) that are becoming increasingly dubious as the intensity and volume of genomic sampling grows and as phylodynamic methods are increasingly being applied at the leading edge of emerging outbreaks and in the face of strong nonlinearities. The project will develop accurate, scalable inference methods with minimal theoretical restrictions, based on recent mathematical advances by the project team. The first builds on mathematical breakthroughs that permit precise estimation of dynamic models from reconstructed phylogenies, while the second seeks to bypass the need for phylogenetic reconstruction altogether by applying new machine learning methods to structured genome-alignment data. Data from the world’s premiere database on avian influenza genomes will be used to resolve outstanding uncertainties regarding transmission within different host species, spillover rates, and seasonality in this system. The work will have applications beyond epidemiology in fields such as systematic biology, cancer biology, microbial ecology, and population genetics. This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.