Most of my work involves either complex networks or mathematical modelling of infectious diseases, or a combination of the two. I am particularly interested in the dynamic interaction of diseases with behavioural, societal, evolutionary and environmental factors. More specific aspects of my research are listed in the following, see also the list of publications for more details.
Modelling Ebola in West AfricaI have been involved in the London School of Hygiene & Tropical Medicine's response to the Ebola outbreak in West Africa. For more information, please visit the Ebola pages of the Centre for the Mathematical Modelling of Infectious Diseases.
Infectious diseases close to eliminationWhen infectious diseases come close to being eliminated, they often behave differently from endemic diseases. Social mixing patterns can shape the dynamics of spread if an infection is concentrated in certain parts of populations. Using mathematical tools, we are trying to understand these patterns and come up with better quantification of the feasibility of elimination.
Vector-borne diseases often have a particularly complex epidemiology, involving the dynamics of the vector population and possible animal reservoirs. The ecological and epidemiological dynamics of both vectors and animal hosts affect the spread of the disease in human populations. An example for this is Gambiense Human African Trypanosomiasis (HAT), a disease transmitted by the tsetse fly and affecting mostly rural populations in Sub-saharan Africa. Gambiense HAT has long been known to infect a variety of animals, but the exact role these animals play in the epidemiology of gambiense HAT, and the effect on disease dynamics in humans, has been a matter of debate. If animal hosts act as reservoirs, that is if they can maintain transmission in the absence of humans, this can have an important role on efforts to eliminate or even eradicate the disease.
Infectious diseases and human behaviour
We often change our behaviour in some way or another when we perceive an infectious disease to be present in our geographical or social proximity. If we avoid contact with those we perceive to be contagious, or when we are ill ourselves, this can change the epidemiology of a disease, and the same can be said for other types of behaviour like vaccination, medication or wearing face masks. Behavioural changes are often triggered by the presence of a disease, but can become independent and be adopted in a population with its own dynamics separate from the dynamics of the disease. Disease awareness, beliefs and behavioural practises can spread in a population through word of mouth or copying others, but can also be influence of mass media or government recommendation. All of these factors generate a complex picture which we need to untangle if we are to understand how diseases spread in populations.
Networks are used as an abstract formulation for a variety of scenarios where individual entities of a greater population interact only some of the other individuals: this can be individual people in a social network connected by friendship or some other kind of relationship, or proteins in a cell connected by their interaction. While networks have an important in visualising such relations, they also allow for the definition and natural interpretation of concepts like clustering and community structure. In network studies, these relations are often assumed as fixed. In many situations, however, links get constantly created and broken up on a continuous basis. I am interested in the question whether any of the structure that is identified in static networks can remain stable even if the evolving nature of the networks is taken into account.