Theoretical immunologist Rob de Boer: ‘Only a small percentage of biologists genuinely enjoy mathematical modelling and programming.’ PHOTO: IVAR PEL, UTRECHT UNIVERSITY
‘The more complex the biology, the more enjoyable it is to model’ – so believes Utrecht-based theoretical biologist and immunologist Rob de Boer, now an honorary member of the Netherlands Society for Theoretical Biology.
‘Theoretical biology is as old as it is very much alive. All four hundred first-year biology students in Utrecht receive lectures on the predator–prey models of Volterra and Lotka, yet we retain at most around forty students who go on to take theoretical biology as an elective. Only a small percentage of biologists genuinely enjoy mathematical modelling and programming. That is a shame, because theoretical biology and bioinformatics have undergone an explosion driven by new developments in biology. We are continually discovering more complex interaction networks, and generating ever more complex data. Biology is an exact science, and I have been arguing for years that advanced mathematics should be compulsory for a biology degree’, says professor of theoretical biology Rob de Boer (born 1958). In his office at Utrecht University, a whiteboard covered in formulae and calculations hangs on the wall. On the bookcase stands a plaque marking his appointment as a lifetime honorary member of the Netherlands Society for Theoretical Biology (NVTB), awarded this summer at the annual NVTB meeting in Schoorl.
By tradition
‘Dutch theoretical biologists, from professor to student, gather there by tradition once a year at the Dopersduin retreat. Many of Paulien Hogeweg’s ‘children’ were there, and she herself was once again simply present’, says De Boer, who this year handed over leadership of his Utrecht research group to Kirsten ten Tusscher. For the time being he continues to work as normal – following the tradition set by his predecessor Paulien Hogeweg (1943), the founding figure of Dutch bioinformatics and computational biology, who, seventeen years after her ‘retirement’, is still present every day. ‘A phenomenon’, De Boer observes. ‘I was one of those Dutch Youth Nature League members who went to study biology to do fieldwork, but became captivated by mathematical models during a course taught by Hogeweg.’
‘The great thing about modelling is that it leads to a deeper understanding of how complex systems work. Immunology is one such complex field, with interactions between hundreds of cell types and various molecules that together decide whether to attack pathogens and to tolerate our own body and our food. That decision is also ‘remembered’ for years in the form of immunity. In such a complex system we seem to drown in data, but models make it possible to move from a purely descriptive approach towards a more functional understanding. Thanks to the modelling of complex data, immunology is also becoming an increasingly quantitative science’, says De Boer.
A turning point
Applying models in immunology did not, however, happen without a struggle. ‘The turning point came in the 1990s with research into AIDS and HIV. From a virological perspective, the infection appeared to be a slow, chronic one. HIV was said to be a slow virus. But using simple mathematical models, American researchers showed that HIV is not a slow virus at all, but has a generation time of at most a few days. That short generation time and rapid evolution explained why HIV escapes antiviral drugs so quickly, and why cocktails of medicines are needed to halt the progression to AIDS.’ The publication of these models – two papers in Nature, 1995 – changed everything, De Boer argues. ‘Before that, you had to go cap in hand to immunologists to ask whether you could test a new model in their laboratory, but afterwards we were actively asked whether we could help interpret their experimental findings using mathematical modelling.’
De Boer focuses primarily on viral infections and the response of immune cells to virally infected cells, in experiments with mice or in human tissue. ‘It is really a matter of straightforward population dynamics. For instance, it is important to quantify how many tumour cells a single immune cell can kill per day. The answer is at most ten or twenty – roughly one per hour – and those immune cells have to contend with a growing tumour consisting of millions of cells’, says De Boer.
‘Mathematical models are always based on grotesque simplifications’
Together with UMC Utrecht, he conducts research using heavy-water labels in mice and humans to estimate the realistic lifespan and production rate of a wide variety of immune cell types. ‘To understand why we remain immune to measles for life, yet only for about a year against influenza or coronavirus, we need at the very least to know how long the cells responsible for this actually live.’
Simplifications
De Boer believes that a theoretical biologist must also be an expert in the biology they are modelling. ‘Mathematical models are always based on grotesque simplifications, and it is quite an art to select the right cross-section from the complex biological reality’, says De Boer. ‘If that fails, a model quickly becomes irrelevant – but because the mathematical model is always included in the publication, it is crystal clear which assumptions have been made and which factors have been left out. Experiments also involve implicit assumptions, but these are often harder to see through. Fortunately, irrelevant models never enjoy a long life and die out quickly.’