Singapore is catching the attention of big IT players Compaq and Cray. The two companies recently established a foothold in the country’s emerging bioinformatics efforts via deals with Singapore’s leading academic institutions.
Cray is collaborating with the National University of Singapore to develop bioinformatics software for whole-genome comparison, while Compaq and Nanyang Technological University are splitting the costs of a S$12.4 million ($6.8 million) supercomputer to support Singapore’s new Bioinformatics Research Center (BIRC).
According to the NTU, BIRC will be “the largest supercomputing facility dedicated to life sciences in Asia Pacific, excluding Japan.”
Under the three-year alliance, Compaq Computer Asia will equip BIRC with an AlphaServer cluster with 64 nodes capable of 512 gigaflops. Compaq expects the system to be fully installed by early 2003.
“Singapore is already renowned for having the best medical training faculty in the region and some of the best computer engineering talents. Now, with BIRC, we will synergize our capabilities by bringing together a community of researchers, scientists, and computer engineers …that will spur Singapore’s growth as a world-class life science hub in Asia Pacific,” said Cham Tao Soon, president of NTU.
Compaq is also sponsoring four scholarships within a new master’s program in bioinformatics at the NTU’s School of Computer Engineering, which hosts BIRC. According to NTU, the program is the first MSc in bioinformatics to be offered by a university in Singapore.
BIRC will be connected to Singapore’s national Biogrid and Compaq will deploy an additional 16 Itanium-based computing units at the NTU as part of the Biogrid project.
Cray is also an active participant in Singapore’s Biogrid effort, and recently unveiled an eight-processor Cray SV1 supercomputer it installed as part of the project at the National University of Singapore Medical Faculty.
According to Cray, the system has given NUS researchers their first access to a supercomputer for bioinformatics applications.
Steve Conway, head of bioinformatics at Cray, said that three NUS bioinformatics staff members have been trained at Cray’s US facilities on programming methods to exploit the “special hardware capabilities” of the Cray SV1 system that were originally designed into Cray supercomputers for use by the intelligence community.
While Conway was mum on the specifics of the project, he said the joint NUS and Cray team is working on whole-genome comparisons between the human genome and “a vertebrate genome.” The team has witnessed a significant speed-up in the comparison process, Conway said, noting that in one example, a problem that took 10 hours to run on another platform required only nine minutes on the SV1. In another example, a project that required two to three days was reduced to three hours, he said.
Conway noted that in addition to speeding up comparative genomics, Cray “is also in the process of porting all the standard bioinformatics applications to the Cray system in such a way that they, too, will be able to exploit these capabilities.”
The most important port will be the Smith-Waterman algorithm, which Conway said is ongoing, but the company also plans to port “the usual suspects,” including Blast, Fasta, ClustalW, and Hmmer.