Research interview – Dr Nik Cunniffe
By Dr Robin Thompson
Robin Thompson talks with Dr Nik Cunniffe, Fellow in Mathematical Biology at the University of Cambridge, about his research in mathematical epidemiology
Your research focuses on mathematical modelling of infectious disease epidemics. How did you get into this field?
After an undergraduate degree in Mathematics at St Catharine’s College, University of Cambridge, I stayed on for an MPhil in Computer Speech and Language Processing. I then worked for a few years, developing clustering and classification algorithms for a search engine company. I returned to university to undertake the MSc in Modern Applications of Mathematics that I believe is still being run at the University of Bath, doing a project at Rothamsted Research with Frank van den Bosch on nematode plant parasites. I enjoyed this experience of research in Mathematical Biology, so continued in this area with a PhD on modelling biological control with Chris Gilligan in the Department of Plant Sciences at the University of Cambridge. I have stayed there since then as a University Lecturer. I also work part-time as a College Lecturer in Natural Sciences for Girton College, where I handle admissions and give mathematics tutorials to biology students, and – although I have given it up now – I was a part-time Associate Lecturer in Mathematics for the Open University for a few years directly after my PhD.
What attracted you to mathematical epidemiology?
In Bath I was taught Mathematical Biology by Jane White, and her lectures initially sparked my interest in applying mathematics to biology and epidemiological modelling more specifically. I consequently chose the MSc project in the area that I mentioned earlier. In mathematical epidemiology I was attracted by the idea of an area with interesting mathematics and a relatively well-developed theory, the possibility of fitting models to data, and plenty of – often very pressing – open questions to work on.
What do you foresee as the biggest challenges in mathematical epidemiology?
I think probably linking models with data. There is lots of very mathematically motivated research in mathematical epidemiology that is incredibly technical and mathematically interesting, but is also quite abstract or very far from a named biological system. In turn this means the work doesn’t necessary help us understand how a particular disease spreads through a population, or what we can potentially do to stop it. Other work concentrates on using quite complicated simulation type models – often including complex risk structures, spatial heterogeneity and stochasticity – to data on pathogen spread. However, techniques for model fitting lag far behind the imagination of modellers. In combination with the often low amounts of data that are available, this constrains how useful the simulation models can be. I think bridging this divide, and working out how to properly parameterise complex models from spread data, is a big challenge.
What is your favourite research paper (by another mathematical biologist)?
I enjoyed James Lloyd Smith’s 2005 Nature paper on the effect of super-spreading on the probability of disease invading; a nice idea, very well analysed.
What are you currently researching?
Pressures of administration and teaching mean I don’t get much time to do my own research any more, and mainly now do research vicariously through my PhD students. Of course, they are all doing wonderful things, in some cases so wonderful I barely understand the mathematics myself. However, I do occasionally undertake some research myself. My most exciting current research work is probably a project that was started during a Working Group at NimBIOS in Tennessee, looking at how interactions between pathogens can be disentangled from survey data. I also have some early modelling results on quantitative (i.e. polygenic) fungicide resistance in plant pathogens.
What impact has your previous research had outside of academia?
A lot of the large-scale simulation modelling work done in the Department of Plant Sciences has been driven by the UK government’s Department for Environment, Food and Rural Affairs (DEFRA), and so has been quite applied. Although I didn’t lead the work, a few years ago I was quite heavily involved in working on models of the likely spread of chalara ash dieback – a fungal disease that affects ash trees – just after it was first detected in the UK. From a standing start, within six weeks we had parameterised models of likely future spread, which were used by DEFRA to support their plan for managing the disease. These models were similar to others I had been working on for some time to predict spread of sudden oak death (an oomycete pathogen from the same genus as potato late blight, which affects a number of forest trees and shrubs). Recent work predicts that the large epidemic in California can now never be controlled; this was highly picked up by the media.
What is your favourite research paper that you have written?
My favourite is probably my first, which I worked on around the sides of my PhD with Alex Cook (who was in Cambridge, but now is at the National University of Singapore). We focused on modelling the effect of point deductions on the chance of football teams being relegated! The main case study is Luton Town – this is where I was brought up – and shows the 30-point deduction applied to Luton Town in the light of financial irregularities was almost certain to lead to relegation the next season, no matter how well they played!
Given the inter-disciplinary nature of your research, you could be in a mathematics or biology department. Why Plant Sciences?
In part it is simply where I ended up, but reflecting on the underlying question, I think I am very lucky to be in a position to work here. As a modeller, I think it is important to interact closely with biologists, since that can lead to new insights and/or topics to work on.
For example, I recently worked with a molecular biologist here to think about the long-term consequences of entirely new experimental results showing that plant pollinators like bees are preferentially attracted to virus-infected plants. We looked at the implications of this using a relatively simple population genetics model; this interaction just would not have happened if I didn’t work here.
I also think it is very important to be in a position to teach basic mathematical biology and modelling to biologists. In Cambridge, we are very fortunate that the first-year biology students almost always studied mathematics in high school. By the end of their first year at university, students are therefore learning reasonably sophisticated modelling; given how biology becomes ever more quantitative, I think this is valuable training for them, and I am very happy to be involved in it.
What advice would you give to a junior mathematical biologist?
Consider working outside the traditional areas of a Mathematics Department; a large number of groups in biology departments are crying out for PDRAs and – particularly – PhD students with strong quantitative skills, but not enough apply. This can lead to closer links with biologists and so interesting problems to work on, as well as nice data to play with.
What is the best part of being a mathematical biologist?
A combination of being able to use interesting mathematical techniques, working with interesting people and having a never-ending range of problems to work on.
Finally, what do you do in your spare time?
I have three young daughters, so – accounting for time spent at work – I don’t really have any!