I first learned about the National Institute of Standards and Technology (NIST) while a graduate student in the Department of Mathematical Sciences at the University of Delaware, which has an active student chapter of the Society of Industrial and Applied Mathematics (SIAM). They regularly invite speakers from government laboratories and industry to give a one-hour seminar about their research and discuss careers outside of academia.
One day, Dr. Anthony Kearsley, a mathematician at NIST, came to give a SIAM lecture about his research. I thoroughly enjoyed the lecture and the applications he described and came away from the talk with an extremely favorable impression of NIST and the research conducted there. In fact, I was so impressed that I approached him after the lecture anxious to learn more about NIST and discuss some specifics regarding his research. After our conversation, he kindly invited me to NIST to spend a day learning more about some of the applications he was working on and to deliver a lecture of my own! I was ecstatic and honored to receive this invitation since my Ph.D. thesis research involved a measurement science problem.
My talk was, fortunately (!), very well-received, and after lengthy discussions, Dr. Kearsley encouraged me to apply for a National Research Council (NRC) Postdoctoral Fellowship. I had first heard about the program from one of Dr. Kearsley’s former NRC postdocs who is a professor at Shippensburg University of Pennsylvania. He recalled his own positive experience at NIST in the NRC program, and he also encouraged me to apply to work with Dr. Kearsley. Looking back as a current NRC postdoc, I agree wholeheartedly with his positive reflections upon his time working with Dr. Kearsley.
In my final year of graduate school, I took Dr. Kearsley’s advice and applied for and was offered an NRC Postdoctoral Fellowship to work at NIST under his supervision. Although I had other opportunities when I got my Ph.D., there was no way I could pass up the chance to work at NIST with such an excellent mathematician and mentor. The NRC Postdoctoral Fellowships are an excellent opportunity that provide recent Ph.D.s with two years of generous funding to pursue their research interests at a unique and vibrant government laboratory. They prepare recent graduates for a wide variety of careers, including those in academia, industry or in other government laboratories like NIST.
One of my favorite parts of being an NRC postdoctoral fellow at NIST is being able to collaborate with world-class scientists on measurement science problems. Such problems naturally lend themselves to interdisciplinary and collaborative work.
For example, I am currently working with NIST engineer Dr. Arvind Balijepalli to construct a mathematical model to optimize the design and maximize the sensitivity of a novel medical diagnostic instrument called a biological field effect transistor (FET). Once fully developed, this device will provide physicians with rapid, accurate and portable measurements of biomarkers—biological molecules whose presence are indicative of a certain disease. Whereas most therapies in use today are prescribed based upon how an “average patient” would respond, physicians can use biomarker measurements to customize treatment protocols for individual patients. Although there are some biomarker tests already being used to deliver personalized therapies, in many cases, making those measurements requires specialized facilities, can take days to weeks to perform, and are very expensive. FETs could make personalized therapies more accessible and affordable.
In addition to my work on FETs, I am also developing machine learning algorithms, a kind of artificial intelligence, for classifying a certain class of drugs known as biologics. Many of the medicines on the market today are small molecule drugs. They have relatively simple structures and are chemically synthesized in a stepwise fashion: add a little of this, a little of that, simmer, strain, bake and voila, you have an anti-cholesterol drug.
By contrast, biologics are derived from living organisms such as insulin (pig and cow pancreases), penicillin (bread mold), and vaccines (most often weakened or dead pathogens). An important type of biologic that has garnered much attention in recent years are monoclonal antibodies. These are therapeutic proteins used to treat diseases ranging from rheumatoid arthritis to certain kinds of cancer.
Many such useful biologics are now reaching patent expiration, which will open up the marketplace to the development of biosimilars, i.e., variants which are clinically identical. Biosimilars must undergo a series of rigorous tests to ensure safety and efficacy before they are released to consumers, so we need to be able to accurately characterize their molecular structure. Research has shown that a technique known as two-dimensional nuclear magnetic resonance (NMR) spectroscopy is an excellent tool for the job.
To facilitate the classification of NMR spectra of monoclonal antibodies, I’m working with my colleagues Dr. Luke Arbogast, Dr. Robert Brinson, Dr. Frank Delaglio and Dr. John Marino at the Institute for Bioscience and Biotechnology Research (IBBR), a joint venture of NIST and the University of Maryland, to develop machine learning tools that can automatically classify different types of NMR spectra. We hope that these tools will ultimately reduce time-to-market for biosimilars and make monoclonal antibodies more accessible and affordable as well.
All the interesting problems and wonderful scientists with whom I get to collaborate make NIST an exciting and stimulating place to be an applied mathematician. Some collaborations involve only one or two people who work very close to my area of research. I also work with larger teams of scientists who are as unfamiliar with my research as I am with theirs, so I’m always learning something new. My research is often aimed at more fully understanding experiments, optimizing experimental design and improving data analysis techniques, so my work doesn’t stop after publishing in a journal or presenting results at a conference.
NIST has a rich history replete with mathematicians whose work deeply influenced the way scientists think about their experiments or interpret their data. A notable example is the work of Dr. Geoffrey McFadden. His work with NIST materials scientists on the solidification of metal alloys has had a profound impact on the field. Dr. Alfred Carasso’s revolutionary work on image deblurring has been used to help scientists interpret pictures taken by the Hubble telescope, MRI and PET brain scans, and helium ion microscope images of nanoscale objects.
My mentor Dr. Kearsley’s work on optimizing mass spectrometers, which identify chemical species by their weight, has changed the way chemists analyze and think about their experiments. More recently, Dr. Paul Patrone, a physicist in my group, has partnered with Dr. Gregory Cooksey, a NIST engineer, and Dr. Kearsley, who is a mathematician, to pioneer novel techniques for measuring the flow of fluid through microscale devices that will substantially reduce both cost and measurement uncertainties. These techniques could eventually be used to help detect cancer. This is a tradition of research in which I am proud to conduct my work and motivates me to conduct the best research I can.