Following the seasonal series theme of “Niche vs. Expansive Research Topics”, I interviewed Dr. Bjarke Frost Nielsen on his journey going from a Physics PhD to working in our EEB department and all of the different topics he’s worked on along the way.
Dr. Nielsen shares, “In general, I have a very broad notion of what physics is. I don’t think for something to qualify as physics it has to, you know, involve Newton’s 2nd Law, be describable in terms of the Schrödinger Equation, or something like that. I think that physics is essentially the science that tries to mathematically tackle the aspects of our physical world that can be attacked mathematically. That’s more or less what physics is, right? It’s choosing the areas where you think that a mathematical description can really capture the problem. … It’s a very broad science in that way.”
Read on to learn more about Dr. Nielsen’s reflections on his research background in Physics and current work in EEB.
Among the various researchers within the EEB department, some may be surprised that a researcher with a Physics PhD is working on epidemiology.
Bjarke Frost Nielsen (BFN): I still very much think of myself as a physicist! It’s actually been a bit of a journey for me, so right now at the moment I mainly study mathematical modeling, that’s kind of the overview, but in a slightly unusual way in that I’m interested in the epidemiology/evolutionary aspect, so how do different evolutionary pressures affect the evolution of a virus and how does that affect the time dynamics of the epidemics that might happen.
During university, Bjarke started his research in quantum physics. Then he worked on relativity, but shifted gears completely to Biophysics when his previous research had unexpected applications.
BFN: Originally, as a master’s student I was doing a degree in Quantum Physics, actually. But I specialized in something that wasn’t technically quantum physics but that was still considered high-energy physics which was essentially the differential geometry of space-time.
But within that framework, I built a model that turned out to have applications for a biological system, which was totally unexpected for me at the time! It turned out that it had some applications for bio-membranes, and I didn’t know anything about bio-membranes at the time but I just thought “Well, that’s cool,” so I looked into that for a bit during my master’s thesis work.
So I started looking more into that and I thought: “Okay, this biophysics stuff is really interesting and I could use some of the stuff I learned in high-energy physics for this!” So that brought me into looking for a PhD in Biophysics.
Then, the research group that I got into at the Niels Bohr Institute was ‘Biocomplexity’, which is really just biological complex systems research. At first I was very interested in all of this stuff having to do with what you would call “soft matter physics,” like membrane physics and that kind of thing. You can imagine “How does your skin develop?” Well, it has to have an outside and an inside, but what’s actually the physical laws governing how that happens?
But the group that I ended up in was doing a lot of work on what you could call kind of “epidemics” in a slightly unusual sense, namely on bacteriophages and bacteria—that’s the viruses that can attack bacteria essentially. They allow you in a controlled, experimental setting to study things equivalent to a pandemic but in a controlled setting, so the evolution of a bacteriophage and so on.
It had occurred to me that there’s a lot that you could use your physics way of thinking for in epidemiology. In areas like statistical physics we are incredibly used to thinking of “How do we model systems of large numbers of interacting entities?” for instance—that’s one of the main goals of statistical physics, in a way. And it’s not so different from modeling interacting individuals in a population and pathogens moving through those populations.
That whole thinking combined with the whole COVID-19 pandemic that broke out in 2020 led me to think that these skills that I’ve been building for awhile, I can see a place where I can apply this. It became very relevant very quickly!
After spending time doing research in epidemiology, Dr. Nielsen found that his physics background gave unique insight into different techniques to solve problems in biology:
BFN: By far the vast majority of epidemiological modeling is this differential equation based modeling—that certainly is very useful and has its place, but is also incredibly coarse-grained in some ways. You’re always treating individuals in large chunks, and saying that “I can describe this with differential equations.”
That’s a good description sometimes, but it hugely oversimplifies some things. It’s sort of saying that each person within that group has the same chance of contacting some person in another group as anyone else. That’s clearly not true, right? I see my wife way more often than most other people do!
You had to—to some extent—do away with the differential equations based modeling and then do something that is very common in some areas of statistical physics which is to use “agent-based” models instead where you model each individual and set up rules for how they interact with other individuals.
That yielded some pretty central insights. It led to an understanding of why lockdowns had been so incredibly successful for something like COVID. [In this model,] reductions in contact network size had really drastic effects on such a disease.
The effectiveness of these methods from outside of traditional epidemiology shows just how crucial interdisciplinary work can be, and Dr. Nielsen notes that he appreciates being from a different background than his colleagues and how they also help further his own understanding.
BFN: Putting those physics methods into the mix helped. That’s one example of something I’ve worked on where physics models make sense. Another is how pathogens spread within the host from cell to cell, through tissue. You can study the rules of how pathogens diffuse and interact with cells in an antiviral state. You can model that from a physics perspective.
It’s a good example of where you need expertise from physics, but most physicists don’t know much about the immune system, which is fantastically complex. So you definitely need outside input from immunologists, for instance. Coming in with some slightly nontraditional methods from physics can really help a problem along.
When asked about working in a more niche field now like the biophysics of epidemiology versus his prior experiences of doing research on expansive topics like quantum mechanics and relativity, Dr. Nielsen noted:
BFN: There’s good and bad sides. I don’t have a huge community of people who do what I do. I can’t go to the “Strings” (String Theory) Conference that has enormous numbers of participants.
On the other hand, it gives me the opportunity to interact with a lot of people from different disciplines: lots of biologists, population geneticists, epidemiologists, immunologists, different people in the medical field, physicists, mathematicians, computer scientists as well.
I’m actually working with Computer Science Professor Adji Bousso Dieng on using some methods from physics and from computer science: entropy-based methods for detecting new viral variants. Which is also combining physics, biology, computer science, data science. It gives me the opportunity to talk with a bunch of different people, which I enjoy a lot!
Interview responses have been lightly edited for clarity and length.
—-
It is so wonderful to see how his journey towards a topic which is quite niche for physicists to do led to close connections with people in fields far different from what he originally researched within physics during university. I thank Bjarke for his willingness to interview and his amazing responses, and I hope that all of our readers enjoy the PCUR seasonal seasonal series!
— Xander Jenkin ‘25, Natural Sciences Correspondent
