Workshop highlights free, fast BLAST processing with friendly interface for faculty and their students
August 13, 2012
Forestry and Natural Resources Professor Andrew DeWoody’s lab focuses on molecular evolution and ecology, including genes involved in evolutionary adaptation.
For doctoral student Nick Marra that means genes from kangaroo rat kidneys, which offer interesting comparison possibilities. Kangaroo rats in the deserts of the southwest United States, for instance, adapted to survive for long periods without drinking water. Meanwhile, their relatives in the Costa Rican rain forest lack the trait. They hardly need it given their home’s plentiful water supply.
Marra’s research also means using BLAST, which has become easier to do recently thanks to a Web-based tool, called BLASTer, running on Purdue’s DiaGrid distributed computing system. DiaGrid and BLASTer can make thousands of processors available at once for BLAST searches and reduce the time needed to set up computations and get results — all at no cost to Purdue users.
Faculty, research staff and graduate students are invited to learn more about and to try BLASTer at a workshop and luncheon from 11:30 a.m. to 1 p.m. Wednesday, Sept. 19, in Room 250 of the Purdue Memorial Union. To register, visit: http://diagrid.org/events/details/5. The event is free, but organizers would like people to register so they can get an accurate lunch count.
Participants are welcome to bring their laptops and input files for sequences they wish to examine. ITaP support staff will be available to assist users with the BLASTer tool. The DiaGrid team also will discuss future plans for BLASTer and collect user input on desired improvements.
Users also can try BLASTer at http://diagrid.hubzero.org/resources/blastgui. BLASTer now supports custom databases, allowing researchers to upload their own protein or nucleotide sequences to search against, as well as data from standard sources such as the National Center for Biotechnology Information.
Marra says BLASTer on DiaGrid has reduced the time for some jobs from days to hours. That gives him time to do more — and more complex — runs to get higher resolution results and to explore what-if scenarios.
“It makes it possible to start tweaking to see what is possible,” Marra says.
ITaP introduced BLAST on DiaGrid in December 2011 and developed BLASTer based on user feedback. The tool is hosted on the DiaGrid hub at http://diagrid.hubzero.org, which is built on Purdue’s HUBzero platform, http://hubzero.org.
“DiaGrid is particularly suited for research using large scale BLAST searches because it involves numerous serial computations that can be parceled out to any number of available processors,” says Carol Song, senior research scientist at the Rosen Center for Advanced Computing, ITaP’s research computing unit. “Generally, the more processors the better, and DiaGrid can make thousands of processors available at a time.”
DiaGrid taps idle processors in offices, student computer labs, cluster supercomputers and more. It is based on Condor, a distributed computing system that works by pooling machines over the Purdue campus network and off campus via the Internet and fast research networks.
Writer: Greg Kline, science and technology writer, Information Technology at Purdue, 765-494-8167 (office), 765-426-8545 (cell), email@example.com