New methods for tracking emergent infections: integrating pathogen genomic and geographic information system data
Lead Research Organisation:
University of Oxford
Department Name: Zoology
Abstract
The rise of emerging infectious diseases during the past 50 years constitutes a major global public health and economic problem, and is due in part to globalization, displacement of indigenous populations and alterations in natural habitats. Understanding the interplay between environment and ecology with pathogen emergence and spread is of great importance to the development of successful intervention and predictive strategies. This research will develop a new set of mathematical and computational methods to determine which ecological variables have the greatest impact on the emergence and spread of pathogens. With this knowledge, public health officials, epidemiologists and governmental agencies can translate the findings into tangible advances to global health. Allocation of resources can then be directed to address factors directly involved in epidemics, such as geographic regions, environmental conditions, or human behaviours.
Technical Summary
The rise of emerging infectious diseases constitutes a major global public health and economic problem. Understanding the reciprocal roles of transmission and environment is critical for developing effective intervention strategies. The spatial spread of pathogens can be modelled within a Bayesian framework known as phylogeography, which incorporates both genetic and geographical data, thereby allowing a priori hypothesis of spatial association to be tested. The complementary field of geographic information systems (GIS) enables testing of positional associations between ecological and landscape variables and disease incidence. Unifying these approaches would allow genomic, epidemiological and geographic data to be analysed in concert. Thus far, however, the use of GIS in molecular epidemiology is sparse, and a rigorous quantitative methodology that integrates the two fields is lacking.
The goal of this project is to design innovative methods that can reveal ecological associations and predictors of infectious disease epidemics. The results can be used to inform health policy, surveillance, prevention, treatment and resource deployment. Specifically, the objective is to integrate phylogeographic and GIS approaches to model the spatio-temporal spread of pathogens by using high-resolution data on geographic and environmental variables. The resulting models will allow statistical testing of correlations between evolutionary, epidemiological, and ecological processes.
To develop and validate this ecological phylogeographic framework, a diverse set of pathogens will be studied, including RNA viruses and bacteria sampled from different geographic regions worldwide. The project consists of three complementary aims:
1) GIS ecological data will be implemented within an existing Bayesian phylogeographic framework. Competing hypotheses can be evaluated by comparing the posterior distributions using model selection tools, such as Bayes factors.
2) Estimates of evolutionary processes (e.g. effective population size and gene flow) will be added to the existing ArcGIS platform, which represents each variable as a separate raster or vector layer. Spatial statistics can then be applied to the various layers to determine correlation among variables, including principal component analysis, factor analysis, and generalized linear models.
3) Ecological variables will be modelled using Bayesian inference using R codes. Evolutionary estimates are used as priors, and interactions between pathogens and a heterogeneous environment formally tested.
As new sequence data becomes available over the course of the project, the models will be tested to determine whether the location/timing of the outbreak was predicted by the new methodology. Methods will be validated using intensively studied epidemics, such as the North American invasion of the West Nile Virus.
The goal of this project is to design innovative methods that can reveal ecological associations and predictors of infectious disease epidemics. The results can be used to inform health policy, surveillance, prevention, treatment and resource deployment. Specifically, the objective is to integrate phylogeographic and GIS approaches to model the spatio-temporal spread of pathogens by using high-resolution data on geographic and environmental variables. The resulting models will allow statistical testing of correlations between evolutionary, epidemiological, and ecological processes.
To develop and validate this ecological phylogeographic framework, a diverse set of pathogens will be studied, including RNA viruses and bacteria sampled from different geographic regions worldwide. The project consists of three complementary aims:
1) GIS ecological data will be implemented within an existing Bayesian phylogeographic framework. Competing hypotheses can be evaluated by comparing the posterior distributions using model selection tools, such as Bayes factors.
2) Estimates of evolutionary processes (e.g. effective population size and gene flow) will be added to the existing ArcGIS platform, which represents each variable as a separate raster or vector layer. Spatial statistics can then be applied to the various layers to determine correlation among variables, including principal component analysis, factor analysis, and generalized linear models.
3) Ecological variables will be modelled using Bayesian inference using R codes. Evolutionary estimates are used as priors, and interactions between pathogens and a heterogeneous environment formally tested.
As new sequence data becomes available over the course of the project, the models will be tested to determine whether the location/timing of the outbreak was predicted by the new methodology. Methods will be validated using intensively studied epidemics, such as the North American invasion of the West Nile Virus.