 |
PAGE 5 OF 6
However, Sunderam and chemistry department chair David Lynn come together regularly as codirectors of the university's Computational and Life Sciences Initiative, launched two years ago to foster interdisciplinary scholarship. They have begun hiring faculty and bringing in graduate students who pursue interdepartmental research in emerging areas that require powerful bioinformatics and other quantitative capabilities. Lynn, an HHMI professor, has also tapped this workforce as a source of teaching skills geared to the new emphasis in biology. For example, he developed a freshman course in which graduate students from a variety of disciplines teach about their research.
Emory's Computational and Life Sciences Initiative has quickly “captured the imagination of a broad spectrum of the community,” says Lynn. “We don't want to weaken the departments, but we do want to catalyze new opportunities between them. That's where the future discoveries will emerge.”
At the beginning of Emory's school year, the students taking introductory chemistry had a choice between a regular lecture format and Morkin's innovative version. Few had ever experienced such an interactive approach to learning, especially in a course considered so fundamental to their future, and “they had to be sold on it,” she recalls.
|
|
|
|
|
 |
|
|
|
|
|
What do math and cancer research have in common? A lot, according to Bert Vogelstein. An HHMI investigator at Johns Hopkins University School of Medicine, Vogelstein searches the human genome for genetic mutations that cause cancer. “There are about 20,000 genes, and each gene has on average 2,000 DNA bases,” he says. “That's about 40 million bases we have to look at” in each tumor cell, and “billions if we're looking at a bunch of tumors.” With all that data, “mathematical analysis is essential,” he says.
Vogelstein says finding cancer-causing mutations among all these bases is especially tough because a mutation doesn't necessarily mean cancer. “Genes are mutating all the time in both normal cells and cancer cells,” he says. Furthermore, the mutations that play a causal role in cancer aren't always the same; most show up in only a small fraction of tumors. The only effective way to identify them, he says, is to use computational tools from the field of bioinformatics.
Vogelstein, who majored in math as an undergraduate at the University of Pennsylvania, calls bioinformatics a way to “distill the wheat from the chaff.” He says the first step is to run computer simulations on DNA sequences from the Human Genome Project to get a sense of how often individual mutations happen due simply to chance. Then, he uses algorithms to compare those mock mutation frequencies with real data from cancer cells. Cancer cells will, on average, have more of some types of mutations than the random baseline created by the simulation, says Vogelstein, and the algorithms flag those mutations. However, not all of them are worthy of further study; some are just “passengers—along for the ride but not actually driving the cancer.” Using researchers' knowledge of gene function, computers prioritize each mutation based on its precise characteristics and the gene in which it occurs. “At the end you end up with a relatively small list of genes” that are worth examining further in laboratory studies, says Vogelstein.
Those studies, which “knock out” genes of interest from cancer cells or the genomes of mice, require huge effort and time. “Before you invest any time in those detailed studies, you need some sort of overall picture,” he says. “The only way to do that right now is to use bioinformatic tools.”
—Benjamin Lester
|
|
|
|
|
|
|
|
|
She showed them studies done at North Carolina State University demonstrating that students who had taken this type of course—so-called “active learning” in a nontraditional classroom setting—earned better grades than their peers in traditional lecture classes. She told them they would need to learn to work together and “figure out the chemistry.” Her enthusiasm—and the data—won them over.
|
|
|
Go Back
