Meet Lecturer James Xenakis in the Harvard Department of Statistics
The Department of Statistics would like to welcome Dr. James Xenakis, Lecturer in Statistics, to our teaching faculty! Dr. Xenakis joined the department in August 2023 after completing a postdoctoral fellowship at the University of North Carolina’s Gillings School of Global Public Health, where he had previously completed a PhD in Biostatistics. In the edited and excerpted interview below, Dr. Xenakis spoke with us about his statistics projects in a mouse lab, reasons for joining the department, and favorite hobby (reading American fiction!).
1. Please introduce yourself (e.g. where you are from, your current work, a hobby that's important to you, and a fun fact about you).
Xenakis: I grew up not very far from Harvard in Concord, Massachusetts and completed my undergraduate degree at that other school in Cambridge (MIT). After college, I worked for about six years at a small pharmaceutical consulting firm, where I specialized in cost-effectiveness modeling of drugs and medical devices.
I’ve spent the last decade or so at the University of North Carolina (UNC) at Chapel Hill. After completing my PhD in Biostatistics there, I took a postdoc position in the Genetics department, where I was the sole statistician in a mouse lab. The lab specializes in breeding inbred strains of mice to help elucidate the genetic underpinnings of diseases. One of my main projects over the last couple years was using mouse resources that we bred to study arsenic-associated type-2 diabetes. Hundreds of millions of people are exposed to unsafe levels of arsenic in their drinking water and, as a result, may also be at elevated risk of developing type-2 diabetes.
My main hobby that’s stuck with me as I've gotten older is reading; no matter how busy or tired I am, I always do a little bit of reading every night. I've amassed a large collection of books, and figuring out where to put them all was a particular challenge in moving from my house in North Carolina to an apartment in Boston. My favorite genre is American fiction, and there's one book that I read almost every year: A Confederacy of Dunces by John Kennedy Toole. This book has an interesting history, too; it didn’t get published until about twenty years after the author’s death, and only due to the persistence of his mother.
A fun fact about me is that I’m a new uncle. My niece was born in April and lives in Massachusetts, which is an extra benefit because now I can look forward to watching her grow up in person rather than on Zoom (even in the past several months, she has changed so much)!
2. Describe some experiences (or mentors) that sparked your interest in statistics.
Xenakis: Statistics appeals to me because I see it as the glue that holds a research project together. For example, in my lab, when a biologist had a research question, the statistician would relate that question to parameters in a statistical model. Then, the technicians on the project would collect the data, and statisticians would become involved again to properly clean, merge, and analyze the data. Since statistics brings together all aspects of the project, a statistician is a natural point of contact for multiple team members with different areas of expertise. The opportunity for collaboration between statisticians and domain scientists always makes the work quite interesting and rewarding.
I had two mentors at my first job out of college at a pharmaceutical consulting firm who sparked my interest in the field and influenced my decision to pursue stats as a career. At the firm, I did cost-effectiveness modeling to support regulatory decisions. I built interfaces that allowed clients to run simulations that were essentially extrapolations and sensitivity analyses (e.g., looking at different model time horizons or populations) based on risk equations developed from their clinical trials. When the lead statistician at the company saw that I was drawn to the equations behind the models, he took me under his wing. He was very talented, trilingual, and probably one of the most prolific readers I’ve ever met (he had the same habit of reading every night). He was a positive role model in that he was incredibly good at statistics but was also an interesting, polymathic type of person.
Another important mentor was my boss and the founder of the company. While my boss was a medical doctor by training, he was also passionate about and quite good at statistics, programming, and economic modeling. To a fault, he would get deeply involved in projects, even at the level of coding (which he probably didn’t have the time for, but he enjoyed it and was skilled at it). From his example, I learned that one of the most important things that statisticians do is to bring statistical thinking, even basic statistical thinking, to a field. This perspective aligns with the ethos of the department here at Harvard.
3. Describe your current research and/or teaching interests. You can discuss a particular project, if you'd like.
Xenakis: As I mentioned earlier, I was the sole statistician in a mouse lab that over the past 20 years has bred approximately 65 new inbred mouse strains. The main highlight of my time in the lab was the collaborative nature of the work; they taught me more science and I taught them some statistics (I hope). My main project in this lab was part of the Superfund Research Program at UNC, which is funded by the NIH to find scientific solutions to protect people’s health, the environment, and communities. I recently helped the lab use these engineered mouse strains to study the genetics of arsenic associated type-2 diabetes.
Working in this lab also sparked an interest in data management, which will be particularly relevant to my teaching this year. I quickly learned that the data generated in a mouse lab are particularly messy, partly because they come from living animals. On top of this, many scientists and technicians use Excel as if it were simply an electronic notebook, instead of thinking about best practices for recording experimental data. My experiences in this lab really helped me understand the challenges scientists face in managing data, and the importance of bringing statistical expertise to a project. It might seem like a humble aspect of our jobs, but you can’t analyze data that isn’t organized and cleaned! I’m interested in addressing the topic of data management in my courses because, while statisticians end up devoting a lot of time to this, it’s also a topic that is often overlooked in our coursework.
4. What made you decide to join Harvard Stats? What are you looking forward to the most for this year?
Xenakis: Hands down, it was the people I met here in the spring that made me decide to join the department. Every time I have an important decision to make, I try to approach it both from an intellectual and an emotional perspective. Often, I find myself having to weigh the two sides against each other. Obviously, Harvard is the “right” choice - this was a unique opportunity to work at such a prestigious institution. But on top of that, being in the department just felt right. Having spent time in many different departments, this one felt uncommonly warm and inviting, with kind and engaging faculty, students, and staff. Fortunately, my logical and emotional sides agreed that joining the department was the best decision to make! In particular, I’m looking forward to learning from the incredibly talented and more experienced faculty this year.