The Committee of Presidents of Statistical Societies (COPSS) annually presents awards to honor statisticians under the age of 41 for their outstanding research contributions and service to further the field of statistics. This year, along with the COPSS president’s award, Nilanjan Chatterjee was selected to receive the COPSS Snedecor award. Chatterjee—known for his important and diverse contributions across biostatistics, epidemiology, and genetics—spoke to Bhramar Mukherjee’s about winning the two awards.

Nilanjan Chatterjee

Nilanjan Chatterjee was born in Kolkata, India, to Aditya Nath and Pranati Chatterjee. He attended Ballygunge Government High School and Saint Xavier’s College before joining the Indian Statistical Institute, Kolkata, from where he obtained bachelor’s and master’s degrees in statistics in 1993 and 1995, respectively. He earned his PhD in statistics in 1999 from the University of Washington (UW) under Norman E. Breslow (biostatistics) and Jon Wellner (statistics). His dissertation, titled “Semiparametric Inference Based on Estimating Equations in Regression Models for Two-Phase Outcome-Dependent Sampling,” won several awards, including the Z.W. Birnbaum award from the UW Department of Statistics and best student paper award from the International Biometric Society, WNAR.

After completing his PhD, Chatterjee joined the biostatistics branch of the Division of Cancer Epidemiology and Genetics (DCEG) at the National Cancer Institute (NCI) as a post-doctoral fellow. He was recruited as a tenure-track investigator at NCI in 2001 and promoted to senior investigator in 2004. In 2008, he was appointed chief of the biostatistics branch.

Chatterjee is known for his important and diverse contributions to biostatistics, epidemiology, and genetics. He has conducted path-breaking research for increasing efficiency of studies of gene-environment/gene-gene interactions, assessing the future yield of modern genome-wide association studies, and modeling subtype heterogeneity for complex diseases. He has made fundamental contributions to theory for the analysis of case-control studies by developing new paradigms that can exploit natural population genetics models for studies of genetic epidemiology. His research in these and other areas is based on many modern and classical disciplines of statistics, including theory of biased sampling, missing data models, semiparametric inference, survival analysis, and shrinkage estimation techniques. He is a prolific collaborator in a variety of scientific studies, including recent genome-wide association studies (GWAS) that have contributed to an understanding of the genetic basis for a variety of cancers.

In 12 years, Chatterjee has published more than 175 articles, many of which have appeared in top-tier statistics and genetics journals. He was elected an ASA Fellow in 2008 and received the Mortimer Spiegelman Award (2010) and Gertrude Cox Award (2011) for his outstanding contributions to the field as a young statistician. This year, along with the COPSS President’s Award, he was selected to receive the COPSS Snedecor award, which is given biannually to a statistician who has made significant contributions to the theory of biometry and had a notable publication within the last three years.

Chatterjee has an extensive record of service as a reviewer and member of editorial boards, award selection committees, scientific program committees, and external advisory boards. Although there is no expectation of teaching at NIH, he has shown commitment to education by organizing and teaching highly attended short courses at statistical meetings. Also, Chatterjee founded and is co-director of the new NCI-Johns Hopkins Biostatistics Predoctoral Training Program. He is a selective, but dedicated, mentor who has trained five postdoctoral and two pre-doctoral fellows, earning himself an outstanding mentoring award from NCI.

In the past two to three years, Chatterjee has advanced both science and statistics by participating as a collaborator and leader in many high-impact scientific studies while maintaining a vigorous methodological research program marked by originality, theoretical rigor, and practical utility.

Following is an interview between Chatterjee and Bhramar Mukherjee.

What was your first reaction to winning the prestigious COPSS President’s Award?

In two words, “Pinch me!” My wife, Barnali, who is also a statistician, and my six-year-old daughter, Chandrima, who has just learned how to use tally marks, were very proud, too. I am humbled that my name will be included among past recipients who are giants in our profession and have inspired many young researchers, including me, for generations. I consider this honor not only a recognition of my contributions, but also of the importance of epidemiology and genetics, two fields I have enjoyed working in for many years.

Which part of your job do you like the most?

At NCI, I am engaged in important scientific problems and I have the freedom and support to pursue those problems that intellectually interest me most. I am happy that, in spite of my many increasing responsibilities, I have been able to preserve some time to sit with a pen and paper and become immersed in formula and Greek notations, whether to statistically formulate a new scientific problem or develop asymptotic theory for a new estimator.

What advice would you give to young people who are entering the profession as PhD students and assistant professors at this time?

For PhD students, I would say learn as much statistical theory and computation as you can. I worry that in the era of bioinformatics and “you-name-it-omics” it may be tempting to skip requirements for statistical theory. It is really our grounding in statistical theory that gives us an edge over researchers from other quantitative fields, such as engineering or computer science, who are increasingly involved in bioinformatics research. It is often easy to pick up the application once you are interested in the field, but learning new theory is difficult. Similarly, statistical computing is a core skill that is best to develop as much as you can when you are a student.

For assistant professors, I would say that whatever you do, it is important to develop a vision so you can identify important problems on your own. One way to achieve this is to take a real interest in scientific applications and develop subject-matter expertise. In-depth knowledge about applications will not only make you a better applied statistician, but also lead you to identify new methodological and theoretical topics of current interest. Further, it is best to avoid posing complex solution of applied problems simply for the purpose of developing rich methods. It is often more productive to come up with simple intuitive solutions that can be grounded with some solid theoretical foundations.

Who are your most significant mentors? How did/do they affect your career?

I have always been very independent when it comes to actual research. But, I have been fortunate to have a number of mentors who have directly or indirectly influenced my philosophy of statistics and science. When I came to the USA from India, I was really interested in pursuing mathematical statistics and probability theory. As a result of a sequence of required courses in applied statistics at the University of Washington—the first of which was taught by my advisor, Norm Breslow—I discovered the appeal of applied statistics. During my PhD dissertation, Norm not only infected me with his passion for research on two-phase study designs, but also taught me about the importance of maintaining a high standard of ethics and fairness in research.

I was able to learn bits of semiparametric and empirical process theory from a series of classes offered by Jon Wellner, which I still find tremendously useful in all my research.

At NCI, I learned from my postdoctoral mentor, Sholom Wacholder, the importance of developing scientific expertise and communication skills to be a successful applied statistician. I received valuable advice, encouragement, and support from a number of other senior investigators, including Mitch Gail, the previous chief of my branch, and Trisha Hartge, a senior epidemiologist in the division. I was also lucky to meet my long-term collaborator, Raymond Carroll, whose deep knowledge of diverse areas of statistics and intoxicating energy for working on new problems continue to inspire me!

How will you compare your experience in education in the United States vs. India?

I have been very fortunate to get the best of both worlds. My wife and I are very proud alumni of the Indian Statistical Institute, where we were able to develop a foundation in mathematical statistics and probability theory thanks to an outstanding set of dedicated professors. The institute has one of the best undergraduate and master’s training programs in statistics in the world and continues to produce some of the brightest PhD students.

My training in applied statistics began at UW, where I really benefited from the close relationship between the statistics and biostatistics departments. What I value most about the U.S. education system is the emphasis on and excitement for applications.

Finally, my training in science and public health began and continues at NCI. I feel very privileged to be at a very happening place, where I can collaborate and learn from so many of the world experts in biostatistics, epidemiology, and genetics.

Why did you choose NCI as opposed to a predictable academic career?

I believe it is important to choose a career based on what you really enjoy, and not on institutional affiliation or job title. I have always been excited by the cutting-edge scientific environment of our division. The unique setting of NIH’s “intramural” research program, which provides funded positions and crucial research support, enticed me to stay at NCI and focus on the research program that I started during my postdoctoral training. In many ways, these jobs are not so different from those in academia in terms of research and mentoring. However, in government, there is a strong collective sense of the public health mission that I have grown to appreciate over time. It is not often fully appreciated that government statisticians—whether working at the National Institutes of Health, U.S. Food and Drug Administration, U.S. Census Bureau, or another agency—often make vital contributions that affect our day-to-day lives. I am very proud to represent government statisticians and certainly humbled to be the first recipient of the award outside traditional academia.

Anything else you would like to share about our profession?

We are living in an era of data. It is a great time to be a statistician or, even better, a statistical scientist. But, don’t take yourself too seriously.

Finally, what are your hobbies/interests beyond statistics?

Yes, I certainly do like many things that do not involve statistics. I like the outdoors generally, and in particular running. I like traveling and exploring new places. And recently, I have developed an interest in playing silly games with my daughter, who just turned six on the day of the award ceremony.