Biomathematics & Statistics Scotland
E. coliO157 is a food-borne pathogen that can cause serious infection in humans: especially the young and the elderly. Scotland in particular, has exhibited some of the highest rates of infection in the world. It is well established that cattle are an important reservoir of infection, with approximately 8% of finishing cattle shedding the organism (1).These bacteria do not cause diseasein mature cattle, and shedding cattle are otherwise asymptomatic.
Until recently, the infection dynamics of this pathogen in cattle were unknown. Mathematical models have been developed within BioSS to describe the within and between animal infection dynamics, assuming that the bacterial population infected only the gastro-intestinal contents (2). These models have proved useful in exploring the likely efficacy of different control methods, both at the in vivo and herd levels, and ultimately have provided evidence of lack of fit when parameterised against data from certain experimentally infected animals. This lack of fit indicates that the observed population dynamics are consistent only with a population which colonises at least one site in the gastro-intestinal tract (GIT). Recently, a site has been identified in the bovine host in which bacterial populations appear to persist (3). It is possible that colonisation also occurs at other, physiologically similar, sites in the GIT.
This project proposes to develop stochastic compartmental and partial differential equation models to incorporate colonisation of the GIT at single and multiple sites, and to explore how the dynamics arising from these models compare with those from the simple models without colonisation, and with data from experimentally infected cattle. The existing between animal infection model could usefully be developed further, to incorporate the properties of shedding cattle as modelled by the in vivo models, as well as to incorporate the recent findings of other investigators, such as those indicating the possible existence of a class of high shedding animals. It is anticipated that a stochastic meta-population model would be a useful addition to the suite of models available to investigate the properties of this infection (4). These models will be used to inform our understanding of current experimental and field survey work, to predict the effect of different control measures on prevalence levels in cattle, and used to explore the properties of different strains of bacteria in the bovine host.
The project will be jointly supervised by Iain McKendrick and one of his university collaborators. Iain McKendrick is currently a member of the IPRAVE consortium (5), an international group of researchers investigating the epidemiology of E. coli O157 in cattle. As such, he is familiar with the data collected within this project, and students will have access to these data. In addition, he is currently a collaborator on a DEFRA-funded project to investigate possible treatments for E. coli O157 (6). This project is generating large volumes of data, creating the opportunity to rigorously parameterise models of bacterial population dynamics.
The student will be based in Edinburgh at BioSS Headquarters on the King's Buildings science campus. Applicants should have good mathematical and statistical knowledge and be interested in the application of statistical methods and mathematical modelling to epidemiological problems.
For further details, contact Iain McKendrick
