Johns Hopkins Tackles the Challenges of Bioinformatics
September 2009--One day Veena Gnanakkan, a third-year Ph.D. student in the human genetics graduate program at Johns Hopkins, discovered that some of her lab mates were trying to align and compare DNA sequences with the human genome using an online database, a process commonly used to identify DNA sequences. Her lab mates were mired in trying to laboriously align 50 or so sequences, one at a time. Gnanakkan, who has a bachelor’s degree in computer engineering and a master’s degree in bioinformatics—and real-life experience in tackling such DNA alignment challenges—informed them of a much faster computational way of doing it.
“Biologists tend to get scared because the method is command-line and requires some coding skills, but it’s really not that complicated,” said Gnanakkan. “And if you have 50 sequences, it’s done instantaneously.”
Analyzing large numbers of DNA sequences and understanding large sets of genes— fairly common problems among biologists these days—often require bioinformatics techniques such as sequence alignment and microarrays, small chips that can contain a whole genome’s worth of DNA, RNA or proteins that are used for large experiments that generate much data. But not everyone at Hopkins is lucky enough to have Gnanakkan in their lab.
Enter Sarah Wheelan. Wheelan joined the faculty at the Johns Hopkins Kimmel Cancer Center in 2006 after earning her M.D. as well as a Ph.D. in human genetics from Hopkins and finishing postdoctoral training at both the National Heart, Lung and Blood Institute and in Jef Boeke’s lab at Hopkins. With her research focused on designing microarrays and analyzing high volumes of sequencing data, she recognized the need for a more collaborative bioinformatics community at Hopkins, one that will provide researchers at all levels with the training necessary to properly analyze their data.
While many such collaborative teams already exist at Hopkins, mostly they arose by chance. And Wheelan believes that with a little encouragement and organization, even more collaborations can be realized.
Along with three other Hopkins faculty members—Jonathan Pevsner of the Kennedy Krieger Institute, Rafael Irizarry of the Bloomberg School of Public Health, and Luigi Marchionni, also of the Kimmel Cancer Center—Wheelan created a proposal that, if funded, would implement new graduate studies, a new research center and new skills development courses in bioinformatics. She proposed that together they would oversee different aspects of the new initiative: Pevsner would focus on curriculum development, Irizarry would guide the research directives, Marchionni would manage the skills development courses, and Wheelan would assist with aspects of all three endeavors.
Last year they submitted their plan in response to a Discovery Request for Proposal from then-provost Christina Johnson and won $200,000 per year for three years, enough money for an upcoming bioinformatics symposium, faculty salary support, computers and publicity about the new opportunities.
“I’m glad the provost felt that this is a really important thing to do now,” said Pevsner. “It’s the right time because there’s a big void here in bioinformatics training.”
The new bioinformatics graduate studies will be, at least initially, more like a training track than a full-fledged graduate program, although no official title currently exists. Students interested in bioinformatics can join existing graduate programs in the school of medicine, such as human genetics or biochemistry, cellular and molecular biology, and complete a thesis project under one or two advisers. For example, students can choose one adviser who focuses predominantly on biological experiments and another who focuses predominantly on computational analysis, and take bioinformatics elective courses to supplement their studies.
Wheelan hopes this new track and students who follow it will help foster new collaborations. Sharing graduate students gets faculty members talking science. A case in point is Gnanakkan. For her thesis work, she designs and uses microarrays to study whether DNA sequences that are repeated in the genome may play a role in diseases like cancer. She started her thesis work with Jef Boeke, who studies repeated DNA sequences. But after frequently consulting Wheelan for help with computational analysis, she took her on as a second thesis adviser. Gnanakkan, Boeke and Wheelan meet for scientific discussions every few weeks or so. Even more frequently, Gnanakkan consults her two thesis advisers separately for research guidance.
“It’s like having a dad and a mom,” said Gnanakkan. “If dad can’t quite help you, you can go to mom, or the other way around.”
For graduate students—both within the new bioinformatics graduate studies and without—and other researchers who have computational problems, the plan includes additional ways to get the appropriate computational help. The Center for Computational Genomics, whose activities are coordinated by Beth Lefebvre, serves as a place where genomics experts can meet with and help basic science researchers with all aspects of large-scale analyses. It temporarily exists in space graciously provided by the Department of Molecular Biology and Genetics in the Preclinical Teaching Building and in a few offices in the school of public health, according to Wheelan. But she and Irizarry note that a virtual space alternative may be sufficient for researchers to get the support they need.
Lastly, the new program includes a series of short courses in biology and computer science. Any researcher interested in acquiring specific knowledge—computer programming or analyzing certain types of data, for example—can take a short course in that skill, likely in the format of a two-hour Web-based tutorial followed by one or two live lectures taught by any faculty member, postdoc or graduate student who has expertise in that particular area. Pilot modules designed by Wheelan, Pevsner, Irizarry and Marchionni will start this September.
Together, the team hopes the new bioinformatics graduate studies, Center for Computational Genomics http://genomics.jhu.edu/, and bioinformatics-related short courses will help create and foster new relationships among a community of like-minded researchers at Hopkins who all need some bioinformatics help. These new changes, however, are not necessarily set in stone: Wheelan wants to try them out and, based on how successful they are, take things from there.
--by Michelle Jones