Research interview – Professor Doron Levy
By Dr. Jennifer Flegg
Jen talks with Professor Doron Levy from the University of Maryland and incoming SMB Board of Directors member, about his research in mathematical biology.
Your research focuses on mathematical oncology. How did you get into this field?
In an orientation event for new faculty, sitting around a round table, we had to introduce ourselves to each other. One of my colleagues at the time, Peter Lee, introduced himself: “I am a haematologist”. I looked back and asked: “What is a haematologist?”. I then introduced myself as a mathematician. Peter Lee looked back and asked: “What is a mathematician?”… This event was the defining moment of my career. It took several years and multiple attempts to start working on mathematical oncology. What started as a side track when I was an Assistant Professor at Stanford University, grew into a full-blown operation down the road, and only strengthened when I moved to the University of Maryland. The University of Maryland is located about 10 miles (16 km) from the white house, or even more importantly, 10 miles from the National Cancer Institute in its main campus in Bethesda, Maryland.
What attracted you to mathematical biology?
My first faculty position was at Stanford University. When hired, I was one of only two assistant professors. Stanford’s math department is a very small department. At the time there were about 20 professors, covering both pure and applied mathematics. There was no expectation from professors in the same department to work with each other. As an applied mathematician trained in numerical analysis, I first considered collaborating with engineers, but quickly realized that my contribution will be rather minimal. As time evolved, I gradually built a reputation of someone who is interested in communicating with other disciplines, and I was approached by different research groups: haematology, radiation oncology, and plant biology. Working in close collaboration with researchers from the medical school and from biology departments, exposed me to the richness of problems that these disciplines offer mathematicians. Yes – as a mathematician I am also fascinated by problems that are still unsolved in spite of decades or even centuries of attempts by top mathematicians to solve them. However, what I find most attractive about mathematical biology is the incredible number of biological and medical problems that have never been studied by the mathematical community. Beyond the novelty of the questions, I am attracted by the level of complexity of biological systems and by the opportunity to bring analytical and quantitative tools to fields that need them so much.
What do you foresee as the biggest challenges in mathematical biology?
Some of the challenges are what I would call “traditional”: communication barriers between mathematicians and biologists; the need to develop interdisciplinary training programs for students (both undergraduate and graduate); strengthening the recognition of the importance of mathematical biology as a discipline within mathematical sciences.
To the previous list, I would add a couple of scientific challenges: The first is the data challenge. While the scientific community seems to be focused on big data questions, I view no data or sparse data as a much more significant scientific challenge. When researchers have a lot of data, at least they have data. My experience is that in most problems I have worked on, there is almost no data. Collecting time-dependent data is critical, yet I have seen so many cases in which patients’ data is collected once a year, so within a period of 5-10 years, it is not uncommon to have 5-10 data points per patient. In many problems we need more data! A second challenge is the future role of mathematical models in a world which is gradually being overtaken by black-box approaches. If we write a mathematical model that includes two dozen parameters, we are likely to face some standard opposition: “how do you estimate the parameters?”, “are you over-fitting the data?”, “how robust is the model?”, etc. In contrast, it seems to be ok these days to propose models that include seventeen million parameters, none of which has any biological or physical meaning. There is value to both approaches, but can mathematical models survive in the age of artificial intelligence and machine learning? Hopefully yes with the aid of some human learning and non-artificial intelligence.
What is something exciting that you are currently working on?
One of the projects I am really excited about these days is a joint work with a postdoc, Heyrim Cho. In a series of recent works we have been combining two fields that have pretty much developed independently: cancer growth and drug resistance. The computational challenges are immense due to the high dimensionality of the problems. This work opens the door to studying optimal strategies for keeping cancer under control. It also provides interesting opportunities to integrate single cell data within continuum models. Quite non-traditional and very exciting.
What impact has your research had outside of academia?
I am a university professor, but every day I am reminded that there is purpose and reason for what we do. And it is not only within the towers of academia. I could have kept my original research program in numerical analysis, but when shifting to mathematical biology I decided to communicate with medical doctors and have them scientifically guide my research. Spending endless hours in hospitals, collaborating with doctors and observing patients, convinced me that mathematical research can be potentially used to improve patient care. It is not only interesting math, it is not only math for the purpose of solving open problems (and there is room for all of that). It is math for solving problems that society cares about and that is a much bigger mission, which I am humbled to be part of.
What is your favourite research paper that you have written?
I would ask the question slightly differently: Is there any research paper of yours you like? Writing papers is always a strange experience. In most cases (but not always) I start writing a paper after I have solved a problem. At that point I may no longer be interested in the problem. After all, it is already solved. Yet, if I had to choose one paper, I would probably pick one of the papers that was co-authored with the “Peters”: my former student Peter Kim and my dear collaborator, Peter Lee (who moved from Stanford to the City of Hope). The paper (published in PLoS Computational Biology in 2008), reported our initial results regarding the role of the autologous immune response in Chronic Myeloid Leukemia. This work received an unusually high level of exposure in the media. It was also the topic of the briefing I delivered on behalf of the American Mathematical Society at the US Congress in Capitol Hill, later that year.
How have you found working with experimentalists?
Challenging (to say the least). You should probably ask the experimentalists how difficult is it to work with me. Generally, I always attempt to work with people that are sufficiently open-minded and are willing to spend time educating me. One of the most interesting experiences I had at Stanford was a week-long training on lab techniques for non-experimentalist faculty. After a full week of working in a bio lab as a math professor, I learned to appreciate the daily challenges that experimentalists face. What I learned over the years is that in order to conduct effective research, it is of utmost importance to continuously communicate with the experimentalists throughout all research phases. In particular, it is key to communicate within the experimental design phase. Otherwise a typical experience is to discover down the road that something that should have been collected or measured, was not.
What is one piece of advice you would give to a junior mathematical biologist?
I will quote Milton Berle: “If opportunity doesn’t knock, build a door.” Do not expect experimentalists and clinicians to be aware of your existence. Don’t be shy. Go and talk to people. Learn the language. Read the textbooks that medical school students read. Participate in seminars in the “application” areas. Make yourself known, and again – don’t be shy. When I was a graduate student I spent about half a year in UCLA as my thesis advisor moved from the Tel Aviv University to UCLA. I rented a room in a Brentwood Hills house that was owned by Alice Beckenbach, a wonderful woman, a widow of two mathematicians: Tucker at Princeton, and Beckenbach at UCLA. Alice told me that she remembered Richard Feynman as a graduate student in Princeton, and that Feynman was the only physics graduate student that regularly attended the Math Department colloquia. Develop your inner Feynman. Don’t ever think that you are too junior to build bridges.
What is the best part of your job? What is the worst part of your job?
The best part of my job is the mentoring aspect. There really is only that much stuff I can do alone as a researcher, but the most satisfying part of my work is the ability to influence science by raising a new generation of scientists. I also enjoy traveling to conferences and developing friendships with colleagues from all over the world. There really is nothing I would consider as the “worst part” of the job. I would probably rather spend less time writing grant proposals, but this is a necessary component of the scientific activity these days, and yes, it does have some positive aspects. After all, writing grant proposals forces you to plan collaborations and organize your ideas. And this is good.
Finally, what do you do in your spare time?
Spare time? I enjoy traveling (and not only to math conferences). I enjoy spending time with my family. Music has always been an important part of my life. I have been playing piano for many years. Don’t be shy about asking me to play the piano if you ever see me next to one. And then there is the other me – “the Magician”. I used to work as a professional magician for many years. I don’t perform regularly anymore, but on special occasions you may experience the magical me. If you ever see me in Los Angeles, I am happy to host you at the Magic Castle, the club of the Academy of Magical Arts. This is the only academy I am a member of.