Nutrition and Stats Grad Students
Mathematician Sanjeena Subedi and nutritional scientist Michael Zulyniak are combining talents to analyze large amounts of data that may help them understand why raiding the refrigerator adds pounds to some people but not others.

Put food and numbers together, and what do you get? At Guelph, it’s a recipe for a new collaboration of researchers in the departments of Human Health and Nutritional Sciences (HHNS) and Mathematics and Statistics who aim to improve human health.

Prof. David Mutch, HHNS, and statistics professor Paul McNicholas both moved here from Europe within the past three years, both enjoy a pint of Guinness and, says Mutch, “We both call soccer ‘football.’”

Now they’ve found common turf in shared research projects at Guelph combining food and stats. Along with a handful of grad students planning joint studies, McNicholas and Mutch are looking for ways to work together and, ultimately, improve human health.

Their meeting place is bioinformatics, the crossover between life sciences and statistics that underlies the University’s new bioinformatics graduate programs.

Bioinformatics involves using computing, math and statistical tools to make sense of biological data. Last fall, U of G introduced two master’s programs and a diploma program, all intended to address a growing need for experts able to analyze mounds of data from scientists studying genes, proteins and other data-dense topics.

McNicholas refines statistical models used to explain patterns in information, including data from one company conducting food sensory experiments and veterinary surveillance data from an Alberta client interested in improving predictive models for diseases such as BSE.

Across Gordon Street, his colleague studies how genes and nutrition interact to cause disease. Health-care practitioners hope to use nutrigenomics to prevent or treat such problems as obesity, diabetes and cardiovascular disease.

Put the two together and you’ve got statisticians hungry for data alongside biologists churning out more information than ever before. Mutch says this is a new link between nutrition and bioinformatics at Guelph.

They’ve begun collaborations through two grad students appointed to their respective departments. Michael Zulyniak began his PhD in nutritional sciences with Mutch early this year; Sanjeena Subedi joined math and stats last fall to start a doctorate with McNicholas funded by the Natural Sciences and Engineering Research Council.

Zulyniak, who grew up in Winnipeg and studied in Saskatchewan and Scotland, says bioinformatics will help in analyzing large amounts of data to learn about metabolic difference between lean and obese people.

That kind of nutrigenomics information might allow doctors to predict and treat health problems, using improved diagnostic tests and diets tailored to individuals or groups. “It’s a brand new field,” says Zulyniak. “I’m really excited about it.”

For her master’s degree in statistics here at Guelph, Subedi looked at models of livestock breeding data for researchers in the Department of Animal and Poultry Science. It was during her Guelph undergrad in biological sciences ― several years after emigrating from Nepal ― that she developed an interest in modelling genetic data. For her PhD, she’ll bring a statistician’s eye to sifting for patterns in nutrigenomics data, perhaps finding genes that work alike or together.

She says her interest reflects the way many statisticians view their role: “The method is there, you just have to learn to apply it.”

Joining the group this fall, Monica Wong is the first master’s student to be jointly supervised by both professors. She will help to generate and analyze data to distinguish the effects of dietary fatty acids on fat metabolism. That work will require her to develop algorithms and apply statistical methods to identify those kinds of differences.

Mutch and McNicholas have also advertised together for another cross-appointed grad student to develop a user-friendly website to analyze gene expression data and enable researchers to predict protein function.