Dr Gianni Lo Iacono is a mathematical modeller at the University of Cambridge working with the Dynamic Drivers of Disease in Africa Consortium, an ESPA– funded research programme designed to deliver much-needed, cutting-edge science on the relationships between ecosystems, zoonoses, health and wellbeing with the objective of moving people out of poverty and promoting social justice.
According to the law of aerodynamics it’s impossible for a bumblebee to fly;
but the bumblebee doesn’t know that, so it flies anyway …
[An old myth, which probably originated as a result of the crude assumptions made by the aerodynamicist who modelled the bumblebee as a static device with fixed wings. An entomologist would have pointed out that bumblebees flap their wings!]
Some time ago, I was looking at a funding body’s policies and came across the word ‘trans-disciplinary’. I am sure that I am not the only one amused by the proliferation of the prefixed-disciplinary family (multi-, inter-, intra-, cross- …). As a mathematical modeller I have worked at the interface between different disciplines for some time and a common question is: how do these whatever-disciplinary teams work?
No doubts there are challenges. Here are some:
- You end up at a conference with a handful of mathematicians surrounded by this magma of life scientists who look at you as a strange animal and test your degree of extroversion by checking if you are looking at their shoes …[*] Alternatively, you are in a seminar in the maths department where everyone is talking about Hilbert spaces and Poincaré Conjecture and you are the only one who has published in Potato Review.
- You need to attend meetings where there is a mathematician, a biologist and a social scientist. It sounds like the opening line of a joke …
- It’s time to publish. But where? Unless you are lucky enough to have your work in Nature or Science, which makes everyone happy, deciding a journal suitable for everyone can be challenging.
Now, I am a strong advocate of the old-fashioned, reductionist approach. Accordingly, any complex scientific problem should be broken down into its basic building blocks. Of course nature doesn’t care about our traditional compartmental division of science and therefore there is no reason to think that the building blocks must belong to one and only one discipline.
Working on a multidisciplinary research programme has reinforced such a view. The programme, the Dynamic Drivers of Disease in Africa Consortium, is considering the links between disease, ecosystems and wellbeing, and doing this by exploring the drivers behind four zoonotic diseases in five African countries, involving research by both natural and social scientists from various disciplines.
Recently, my work on this took me to Sierra Leone to discuss our field work on Lassa fever. Lassa fever is a rodent-borne viral disease. It is common in West Africa and the Lassa virus infections per year are estimated at 100,000 to 300,000, with approximately 5,000 deaths. Because of the high incidence and the relative abundance of data, Lassa fever is an ideal candidate to model rodent-borne diseases. It also presents interesting features, such as seasonality in the rodent abundance, seasonality in disease prevalence and, perhaps, seasonality in human incidence.
As a mathematical modeller/physicist this sounds great. Surely the seasonality in human incidence must mirror, or somehow relate to, the periodic fluctuations in the disease and in the number of rodents? Then you start to think about suitable methodologies. The standard tools of time-series analysis should do a good job: you’ll find at which frequency cases are reported, compare this with the frequency of abundance and prevalence, and finally try to infer the underlying mechanism. A simple analysis and job done!
Until you talk with Dr. Donald Grant, director of Sierra Leone’s Kenema Government Hospital which specialises in Lassa fever, who tells you that, yes, in the past they have observed some degree of seasonality in human incidence, but now people go to the hospital any time of the year.
This shouldn’t happen; it messes up my simple time-series analysis. Did the rodents decide to get sick in the wrong season? So you start to ask your team more and more questions.
The social scientist tells you that people often go to the local healer for help and therefore many cases are not reported. (So maybe I need to include some sort of censoring in my simple analysis?)
Then you observe people of the village during the participatory modelling being undertaken, and one outcome is that they no longer eat the rodents because they have been told that this causes Lassa. (So people’s behaviour has changed. Do I need to allow for some trend in my simple analysis?)
While you wander in the village you notice that some houses are built using different materials. The local people and the environmental team confirm that the microclimate (temperature and humidity) is indeed different in these houses, and the epidemiologists and the biologists explain that the stability of the virus might depend on the local temperature and humidity. (So in my, no longer simple, analysis I need to include this additional effect.) While the discussion with your team goes on, new questions emerge, your model changes, the design of the experiment is re-shaped …
Visiting Sierra Leone was an amazing experience and although I came back with more questions than answers I learned a lot from my colleagues and from the local people. At the beginning of the project my priority was doing good science and producing good papers relevant to my discipline. Of course I still hope that will be so.
But I think that doing something useful would be a great (perhaps the greatest) achievement too. And I know more than ever that is only possible by considering the multitude of resources and processes, as well as their strong mutual interactions, behind the ecosystems we are considering. The trip to Sierra Leone illustrated clearly that forcing ourselves to allocate each building block of a complex ecosystem to one discipline alone seems only to set up a path for failure.
So whatever the challenges to multidisciplinary working – just like the many jokes aimed at us mathematical modellers – I think they are worth overcoming.
What’s the difference between an introvert and an extrovert mathematician? An introvert mathematician looks at his shoes when he’s talking to you. An extrovert mathematician looks at YOUR shoes.