Mathematical Modelers

in applied health sciences compare notes, share techniques

By Laura J. Bernhard

For the seventh time in 15 years, a group of mathematical modelers in the nutrition and health sciences met to exchange ideas and compare techniques during Mathematical Modeling in Nutrition and the Health Sciences, held at The Penn Stater Conference Center Hotel. This conference brought modelers from around the world together for three days and nights of lectures, discussions, paper presentations, instructional talks, data modeling workshops and software demonstrations.

Mathematical modeling is the application of mathematical and statistical techniques to the investigation of biological systems. A model, derived from data, can be used to virtually simulate experiments and test hypotheses, which can result in better research. According to Dr. Michael Green, Penn State professor of nutrition science and physiology, most modelers are self-taught. While Penn State offers a graduate course in mathematical modeling, taught by Green, it is still extremely rare to find such an offering. Previous meetings of this group of scholars have been held at six different university locations throughout the United States. Eighty-five professional modelers and students attended the Penn State conference.

Conference organizers were Green; Dr. Ray Boston, professor of biomathematics and biostatistics at the University of Pennsylvania School of Veterinary Medicine; and Dr. Janet Novotny, research physiologist at the Diet and Human Performance Laboratory of the U.S. Department of Agriculture.

Keynote speaker and honored guest Dr. Britton Chance, Eldridge Reeves Johnson University Professor Emeritus of Biophysics, Physical Chemistry and Radiologic Physics at the University of Pennsylvania, set the tone for the conference with his opening address. Chance is a pioneer in mathematical modeling and formulated many of the accepted ideas and technologies in the field.

Computer capacity is critical to being able to successfully model. As computer technology has been enhanced, a scientist’s ability to model quickly and effectively has increased. Equations that took hours to process in 1960 can now be solved in fractions of seconds, complete with graphical user interfaces.

The conference featured evening data modeling workshops where various software systems were demonstrated. Participants had the opportunity to work data sets prepared by the facilitators and also had access to the computer and modeling experts to experience the software for use on their own data. According to Green, “It is more and more essential at these meetings to have computer labs set up.” At these labs, it was possible for participants to “see new aspects of the software and spend time working with each other’s data.”

The conference provided sessions geared to both experienced and student modelers. Instructional talks for new modelers were presented by Dr. Judah Rosenblatt, professor of biostatistics at the University of Texas, Medical Branch, who spoke on “Converting a Physiological Account of an Experiment on a System to a Model Portraying, in a Mathematical Sense, What Transpires”; Dr. Meryl Wastney, chief scientist, Metabolic Modeling Services, Dalesford, Hamilton, New Zealand, who discussed “Fitting a Mathematical Model to Biological Data”; and Dr. James Hargrove, associate professor of foods and nutrition at the University of Georgia, who talked about “Interpreting the Outcome of a Model-building and Model-fitting Exercise.”

Experienced modelers presented papers on the following subjects: the utilization and metabolism of vitamins and minerals, the modeling of lipid soluble compounds, and protein metabolism. Dr. A. Daniel Jones, Penn State’s director of mass spectrometry, discussed “Advances in Mass Spectrometry” as part of the conference session on the experimental aspects of kinetic data.

The conference ended with paper presentations on modeling energy, growth and development and modeling to explore and assist with understanding disease processes. Boston announced the AKA-Glucose Project, a joint project that will shape the field of modeling in the years to come. The conference also featured a poster session, with 10 poster presentations on subjects such as the “Modeling of Dietary Intake Distributions” and “A Matrix Model of Fasting Metabolism in Northern Elephant Seal Pups.”

Joanne Green, research assistant in nutrition at Penn State, is editing the conference monograph, which will be forthcoming. The program was sponsored in part by Agribrands International; Hoffmann-LaRoche; the National Institutes of Health; Purina Mills Inc.; Penn State’s Department of Nutrition, the College of Health and Human Development and the Life Sciences Consortium; the U.S. Department of Agriculture; and the University of Pennsylvania.

An outreach program of the College of Health and Human Development and the Department of Nutrition

Dr. Janet Novotny (center), research physiologist at the Diet and Human Performance Laboratory of the U.S. Department of Agriculture, presents information on the computer software WinSAAM, used for mathematical modeling, during the Mathematical Modeling in Nutrition and the Health Sciences conference.

Dr. Ray Boston (right), professor of biomathematics and biostatistics at the University of Pennsylvania School of Veterinary Medicine, demonstrates the mathematical modeling software STATA to participants during the Mathematical Modeling in Nutrition and the Health Sciences conference, held at The Penn Stater Conference Center Hotel.

Dr. Michael Green (left), Penn State professor of nutrition science and physiology, talks with Dave Polidori of Entelos during the poster session of the Mathematical Modeling in Nutrition and the Health Sciences conference. This was the seventh meeting for mathematical modelers working in applied health sciences.