- HPC clusters help analyze tens of thousands of drosophila genes.
- Purdue Synder cluster is set up specifically for life science research.
- Analysis is answering questions about aging and deteriorating eye performance.
Vikki Weake’s lab has its eyes on the genes involved in sight — and sight’s deterioration with aging — research that might lead to new ways of prolonging the eyes’ lifespan.
“The eye is actually very accessible for treatment,” notes Weake, an assistant professor of biochemistry at Purdue University.
Her research involves sorting through changes in tens of thousands of genes from aging fruit flies, the lab’s model organism, and millions of pieces of sequencing data from those genes.
“What we want to do is look at genes in the population at the transcript level over time, so you can imagine the number that we're looking at is fairly huge,” Weake says. “That's why we have to use the computing clusters. You can't do it otherwise. Just to figure out what things are changing and what things are not changing, we've got to analyze it on a big scale.”
ITaP built the Snyder cluster in 2015 at the same time it installed the new Rice cluster. While Rice is geared to physical sciences and engineering research, Snyder is designed for life sciences, including large memory capacity and appropriate software array, among other things.
Weake’s lab uses edgeR, a version of the R statistical computing package available on Purdue’s Community Clusters, together with complex statistical models developed by Weake’s collaborator Rebecca Doerge, Trent and Judith Anderson distinguished professor of statistics. The changes they look for are subtle and require powerful statistics to identify, another reason the research is computationally demanding.
“The big question is why do some genes change over aging and other genes don't, and what happens first,” Weake says. “Because if you could understand what happens first, that's the part you could target to prevent age-related changes. If you could reverse it at the beginning, you could prevent it.”
Fruit flies have distinct advantages as a model organism. Their genome is well characterized, they’re readily available and they grow old in less than a couple months. Nature also has reused — or conserved — many genes, especially disease-associated genes, and many basic genetic processes among fruit flies, humans and other organisms.
“We'll need to test and see if the same mechanisms are conserved in humans and that would be the next phase,” Weake says. “But looking at fruit flies allows us to identify mechanisms much more quickly and to test our hypotheses in a way that we can't test them in human eyes.”
Access to the Snyder cluster through the department of Biochemistry makes Weake’s computational work convenient and comes with support from ITaP Research Computing’s expert staff, even to the point of working with her students in the lab to get jobs up and running.
“The department having some nodes is actually fantastic because there's a lot of people doing these type of analyses here,” Weake says. “Graduate students 10 years ago didn't do this. Now, many of the graduate students have to learn some programming, have to learn how to analyze their data so they can do it themselves. It's part of the growing toolkit of a graduate student in biochemistry.”