Broadband engine speeds discovery

CHELMSFORD, Mass.—Mercury Computer Systems Inc. and Boston University have success­fully migrated a fragment-based drug design (FBDD) application to the Cell Broadband Engine (BE) processor, achieving an order of magnitude improve­ment in processing time in a smaller system footprint over previous configurations.

 

FBDD is a promising new approach in the pharmaceutical discovery and design industry that depends heavily on com­puter simulation. FBDD simu­lates the chemistry and phys­ics of molecular interactions in order to estimate how well potential drugs will bind to their target proteins. The joint goal of the Mercury-BU collaboration is to leverage the team's experi­ence and know-how to develop innovative, commercially viable drug discovery products that, unlike similar methods used both in academia and industry, don't require X-ray crystallog­raphy, NMR or other supportive methodologies.

 

"This approach has the poten­tial to revolutionize the cost and pace of new drug development," says Dr. Sandor Vajda, profes­sor of biomedical engineering at BU. "With Mercury's hardware, software, and assistance in algo­rithm optimization, this method is more commercially viable."

 

The Structural Bioinformatics Lab of BU initially developed a highly regarded FBDD tool that creates a map of likely drug bind­ing sites on the surface of pro­teins. However, this program took weeks to run on a departmental Linux cluster. Although later software iterations led to signifi­cant improvements on the Linux cluster, and even more dramatic improvements running on an IBM Blue Gene cluster, neither rivaled the processing speeds achieved on Mercury's Cell BE processor-based hardware.

 

The team successfully migrated the FBDD computer simulation in progressive steps from a shared departmental Linux cluster run­ning for weeks to a single Cell BE processor running for less than three minutes. Mercury and BU report that the average computa­tion time for the application run­ning on the Cell BE processor is approximately 10 times faster than the same application running on a BlueGene processor, in a chip-to-chip comparison. This enhanced speed will enable biotech firms to use the newly created algorithms as a small molecule discovery and design platform.

 

"Moving away from a shared supercomputing infrastructure on a cluster to a dedicated supercom­puter on a single Cell BE processor can make a tremendous difference in productivity for a development team," said Mirza Cifric, director of the Biotech group at Mercury Computer Systems.

 

Clusters are traditionally col­lections of otherwise independent computers, such as Pentium work­stations or rack-mounted PCs. "Conversely, multi-core processors such as Cell BE are a collection of processing element that share the same silicon providing, in effect, multiple PCs on one chip," Cifric explains.