Dr Anthony Wood
Postdoctoral Research Fellow
Background
I am an epidemiologist and data scientist in the Kao Group, working on models of infectious diseases. Before joining the Roslin, I completed my PhD in the School of Physics and Astronomy. This was in Statistical Physics; I looked at mathematical models of non-equilibrium systems. Between the PhD and this current role, I spent time in industry.
Bovine Tuberculosis (bTB)
bTB is a disease that has profound impacts on global livestock, both economically and socially. Control of the disease is often complicated by the presence of additional wildlife hosts; in the case of Great Britain, this is predominantly badgers. It is difficult to disentangle the roles of different species in the overall disease system, however, but doing so is crucial for making informed decisions about control measures. My work looks at how best to use different strands of data (host life histories, movements, pathogen whole-genome sequences) gathered across both badgers and cattle in GB to infer the roles of these different species. This is both by calculating the likelihood of transmission from one host to another based on all the data we have on the hosts, and more recently in developing more sophisticated "target metrics" for fitting (ABC-SMC) models of bTB transmission, to better identify the roles of specific hosts.
This is a collaboration with Biomathematics and Statistics Scotland.
COVID-19
I developed machine learning statistical models to explain spatial and socioeconomic variation in COVID-19 cases, severe outcomes and vaccination uptake across Scotland. These variations were constantly evolving with changing non-pharmaceutical interventions policy, and new variants of concern. My primary focus was on understanding the role of deprivation and how that changed over time, and I regularly contributed to SPI-M-O/Scottish Government/Public Health Scotland modelling efforts during the acute stages of the pandemic.
Wastewater-based epidemiology (WWBE)
WWBE is emerging as a new means of monitoring the presence/absence or prevalence of infectious diseases in a population, without individual testing. Using data on real wastewater systems in Scotland and representing those systems as a tree-like network, I am developing methods for best placing monitoring sites to optimise the system's effectiveness (i.e. how many infections are there before it is detected by the system), bearing in mind that sampling may not be daily, and the samplers do not have perfect sensitivity.