Combined Efforts: Bioinformatics at Vassar

By Jeff Kosmacher

Biology 353a. Bioinformatics (Same as Computer Science 353) DNA is the blueprint of life. Although composed of only four nucleotide “letters” (A, C, T, G), the order and arrangement of these letters in a genome gives rise to the diversity of life on earth. Thousands of genomes have been partially sequenced, representing billions of nucleotides. How can we search this vast expanse of sequence data to find patterns that provide answers to ecological, evolutionary, agricultural, and biomedical questions? Bioinformatics applies high-performance computing to discovering patterns in large sequence datasets. In this class students from biology and computer science work together to formulate interesting biological questions and to design algorithms and computational experiments to answer them. Ms. Schwarz and Mr. Smith.

Biologist Jodi Schwarz came to Vassar in fall 2006 fresh from the U.S. Department of Energy’s Joint Genome Institute (JGI), one of the centers where the breakthrough sequencing of the human genome was accomplished in the late 1990s and early 2000s. Schwarz’s fellowship at JGI provided an invaluable lesson for all twenty-first century scientists: exceed conventional boundaries.

Half of her coworkers on the sizeable staff were computer scientists, and Schwarz quickly learned how critical it is for a biologist to know advanced computational approaches. “You can’t study high-level genome questions without them,” she says.

Comfort and skill with computer science methodologies are becoming essential in the emerging world of computational science, where researchers apply the analytical savvy of computer science to model scientific phenomena, and where a computer provides the laboratory for experimentation. Epitomized by the Human Genome Project, computational science is now rapidly expanding throughout the natural sciences, including chemistry and physics.

“In computer science we learn searching and patterning techniques to find a ‘killer app[lication]’ that will make large amounts of information meaningful and useful,” says Marc Smith, an assistant professor of computer science who is partnering with Schwarz. “As we began to apply to biology the searching techniques we initially used for text or language, the amount of data to analyze grew from shorter sequences to the human genome of over three billion nucleotides. That’s no longer a project you can apply elbow grease to. You really need a computer program.”

Vassar hired Smith and Schwarz to bring their departments together around the biology-focused computational science known as bioinformatics. Spawned by the genome sequencing revolution, bioinformatics combines mathematical, statistical, and computational methods to model and analyze biological systems. Their partnership is key to Vassar’s success in advancing computational science in the curriculum over the last decade.

Nancy Ide, professor of computer science and department chair, helped set the stage for bioinformatics when she convened meetings of the various science chairs to explore the curricular possibilities for computational science within their disciplines — from biology to chemistry to neuroscience to physics.

In 2000, Vassar first presented its vision to broadly embrace bioinformatics in a successful grant proposal to the Howard Hughes Medical Institute, which awarded the college $1.3 million. The multifaceted proposal — written by longtime biology professor J. William “Bill” Straus — sought support for a biologist (eventually Jodi Schwarz), who would specialize in genomics and bioinformatics, and also collaborate with the chemistry department. Another lynchpin was an upper-level course to be co-taught by that biology professor and a computer science colleague to students from both departments.

As Schwarz and Smith designed the new bioinformatics course, they had one immediate challenge: “Traditionally, biology students are very afraid of high-level computation,” explains Schwarz. Also, at the Joint Genome Institute she experienced communication gaps between computer scientists and biologists, sometimes the result of the two disciplines using the same terminology to say different things. Ultimately, the professors agreed that their primary task would be to teach young science students to think and communicate differently and to comfortably cross the intellectual lines of disciplines. As it turned out, Smith and Schwarz became their own best example.

“Working with Marc has shown me different ways to approach teaching and problem solving,” says Schwarz. “He has insights that I would never have because his expertise and his teaching style are so different. Even so, our styles and perspectives are very complementary, and the students learn a lot just from watching us interact!"

Marc Smith and Jodi Schwarz
Marc Smith and Jodi Schwarz

The professors' collaborative style is an inspiration to students.

Smith earned his computer science Ph.D. while at AT&T, where for fifteen years his work ranged from IBM mainframe and Unix applications to systems engineering. Even during his subsequent five years teaching at Colby College, he had primarily explored bioinformatics with other computer scientists. Schwarz has a Ph.D. in zoology, and before joining the Vassar faculty she had completed post-doctoral fellowships in parasitology and genomics.

Schwarz and Smith started at Vassar in the same semester, and almost immediately began an intensive tutorial with one another on their disciplines, meeting once or more a week. Within a year they essentially translated the lessons of that shared learning experience into Bioinformatics 353/353a (described in the course catalogue excerpt in the first paragraph).

Biology major Nooreen Rubin ’09 took the course when it was first offered, but not without some reservation. “I could surf the Internet and make PowerPoint presentations but I knew nothing about computer science,” she recalls. “The semester before, I had used computers in the genomics class that Jodi Schwarz teaches. We had done a lot of BLAST [Basic Local Alignment Search Tool] searches, where you take a DNA sequence and compare it to millions of other sequences. I got a result and I understood the data, but my mind was still so geared toward biology.”

“In Bioinformatics,” Rubin continues, “we looked at the scripts for how a BLAST search works. Being able to write a script that analyzed the DNA I’d been studying, that’s when it clicked for me. And then we wrote other scripts to pull out the tiny bits of information we needed from huge BLAST searches. It made so much more sense and I got so much more data out of it,” says Rubin.

Right after completing the course, Rubin landed an internship last summer at a Cornell University genetics lab, where she confirmed the power of her scientific evolution. “I was working on a project with a lab technician who said, ‘Hey, I’ll write the script,’ but before I knew it I had shown him some tricks I’d learned to make the script better,” she says.

Helping students to recognize and build upon their growing scientific expertise is key to how Smith and Schwarz approach their course, which emphasizes original research projects. So when the bioinformatics professors heard that chemistry professor Teresa Garrett had hit a stumbling block in her research, they suggested their class might be able to help.

“As a trained lipid biochemist, I have some rudimentary skills with bioinformatics,” says Garrett, an associate professor of chemistry. “But while the search that I wanted to do was completely beyond my skill set, it was quite straightforward to the students. Their ability to tackle this jumpstarted an important process for me. They had skills that I didn’t and those skills would facilitate my research program. I think that was the reason they were so jazzed about the whole thing.”

Computer science major Philip Tully ’10, biochemistry major Samuel Woodle ’10, and biology major Amory Meltzer ’09 were so inspired by this experience that they launched  Bioinformatics Think Tank (BiTT) the following semester, presenting themselves to the Vassar faculty as a research support team and recruiting fellow students to join their start-up. That BiTT’s services were quickly in demand is a powerful reflection of the hands-on learning that Smith and Schwarz emphasize in their course. (Emboldened by their growing confidence, the BiTT cofounders have also reached beyond Vassar’s walls and approached the National Cancer Institute about collaborating on a complex problem related to RNA folding.)

Jodi Schwarz
Jodi Schwarz

Schwarz, back, with biology graduate Rubin and BiTT cofounders Woodle and Meltzer

Another of Schwarz and Smith’s teaching strategies is to pair students from different scientific backgrounds on the experiments — the course has enrolled not only biology and computer science students, but biochemistry, cognitive science, and neuroscience and behavior majors as well. “We want our students to collaborate scientifically and to be both teachers and learners,” Smith points out. “When they have to explain what they know, it helps them to better understand the material and how to communicate effectively with people who have different scientific perspectives.”

“I never felt more excited to do work than in that class,” says Philip Tully. “I wouldn’t have even glanced at the biology course offerings if it weren’t for the bioinformatics course. Now that I’ve also completed 106 (Introduction to Biological Investigation), I may expand into courses like Genomics and Molecular Biology. What’s really cool is if you had asked me a year ago what I would be doing now, I never would have thought this would be in my future.”

Like his biology counterpart Nooreen Rubin, Tully’s immediate future included a prestigious internship. This past summer at the Cold Spring Harbor Laboratory in Long Island, New York, he put his bioinformatics know-how to work for the new iPlant international plant sciences consortium.

Cellular and molecular biologist Bill Straus, the principal investigator for Vassar’s pivotal Howard Hughes Medical Institute grant, notes that, “A challenge in the sciences, especially biology, is keeping up with the incredibly rapid accumulation of knowledge and evolving technology.”

He explains, “Bioinformatics was an obvious need in the biology curriculum, and wasn’t a part of the our faculty’s communal experience, so the department quickly agreed to make it a focus of the proposal. We also knew that the computer science department was very excited about jointly developing a bioinformatics curriculum,” Straus recounts.

Later, when Smith and Schwarz needed $35,000 for a computing system powerful enough to support their new venture, Vassar allocated funds from the Office of the Dean of the Faculty, the Frances Fergusson Technology Fund, and the biology and computer science departments.

This curricular strategy helped move Vassar into the thick of the liberal arts colleges leading undergraduate computational science education. In 2007 Vassar, Barnard, Carleton, and Williams colleges, along with Columbia University, secured a three-year grant from the Teagle Foundation to advance genomics studies at small colleges. Later that year Vassar hosted the first in a continuing annual series of workshops to help faculty bring hands-on genomics research into the classroom.

While Vassar has had the foresight in the past decade to advance bioinformatics in the curriculum, it is hardly a newcomer to computational science. Such notions had begun to take shape in Vassar’s nascent computer science curriculum — one of the first of its kind at a liberal arts college — ushered in by mathematics professor Winifred “Tim” Asprey in 1963. Marc Smith is pleased that he had a chance to tell Asprey about the bioinformatics initiative, shortly before her death in 2007. “She was so excited,” he recalls.

That legacy and Vassar’s strong history in multidisciplinary learning have provided Smith and Schwarz with a powerful foundation. What’s the most important thing the teachers have learned so far from their experience?

“I have really appreciated how much more we can accomplish as a team,” says Schwarz. “The complete respect and trust Marc and I have for each other’s expertise and teaching instincts has been a very special experience. Even though it requires so much more effort to teach as a team, the rewards are enormous and something I never could have imagined before our course.”

Teamwork: Add that to the “must” list for twenty-first century scientists.

Jeff Kosmacher is Vassar’s director of media relations and co-chairs the college’s Campus-Community Advisory Committee.

Photo Credits: Spencer Ainsley
Have comments about this article? Email vq@vassar.edu