Using mathematical modelling and statistical approaches to fit models to data, I work to gain insights into a broad range of disease problems affecting farmed livestock. Modelling is a useful tool that allows us to represent complex systems in a computer, and statistical tools allow us to 'tune' those models so that they give us the closest possible representation of any data we have available. 'Mechanistic' models allow us to incorporate what we know about how a system works, insights that aren't always in the form of data but are central to our ability to make predictions. Using these tools, we can explore how to deal with specific disease threats. Much of my work has focussed on Johne's disease, a gastro-enteritis that causes farm animals to stop being able to effectively digest food and gain weight. I've also worked on several other systems, including bovine tuberculosis, fish diseases and forestry pests. 

While modelling is limited in how detailed and accurate a representation of a system it can achieve, it is important to understand that insights gained from them often hold true in the real system. Imagine for example comparing two models of disease transmission, such as aerosol transmission vs transmission arising from shedding onto pasture. These different routes will produce different patterns and we can use our models to discern which is likely to be playing the decisive role. This in turn gives us leverage in designing approaches to prevent disease, for example if vaccines are available which animals should be prioritised to have the most impact.