YouTube Facebook LinkedIn Google+ Twitter Xingrss  



By Salvatore Salamone

June 12, 2002 | Life science companies want more from distributed systems than just raw computational processing power, and a partnership between Platform Computing Inc. and SciTegic Inc. promises to address that need.

Specifically, Platform is combining its workload management software comprised of the Platform LSF and Platform JobScheduler, with SciTegic's Pipeline Pilot, a high-throughput data analysis and mining system for drug discovery informatics. The combination, called the Platform Life Sciences Suite, is a data pipelining solution that will accelerate time to discovery, according to the two companies.

“Drug discovery data is being grossly underutilized due to the combination of increasing volumes and types of data, and the dependence on computational methods,” said Yury Rozenman, Platform’s director of life science business development.

The Platform Life Sciences Suite should help make some data pipelining situations easier because the software can give a manager information about the computation resources available. Additionally, the software can break data analysis or mining pipelining tasks into discrete components, each sent to run on different computers in the distributed system. The Life Sciences Suite also tracks the dependencies among individual execution steps. In this way, discrete jobs are performed in an appropriate order.

Such capabilities aimed at the drug discovery process are in demand. Companies want to use distributed computing systems to improve the efficiency of the entire drug discovery process.

For example, earlier this year Novartis Pharmaceuticals started using a distributed computing system from United Devices Inc. to speed its in silico research efforts. Novartis wants to accelerate lead identification and profiling to increase the number of disease relevant targets tenfold in the next year.

Clustered and distributed computing are a core component of the IT strategy of virtually all medium- and large-sized life science companies, according to the market research firm Silico Research Ltd.  A market study released earlier this year by the Silico group found that 85 percent of large pharmaceutical companies and 65 percent of other life science organizations use distributed computing systems.

Once a distributed system is in place and managers see its value for one application, other uses for the system come to mind. This is what is happening today with distributed computing as it relates to drug discovery.

“This pattern is quite common when a new networking technology is deployed,” says Raymond Lopez, an independent systems consultant. Lopez notes that over the last few years many companies have installed secure collaborative systems for a narrowly defined purpose (collaboration within a department). And after managers saw the benefits, the systems were opened up to other applications, departments, and outsiders.

“The same thing is happening today with distributed systems,” says Lopez. “You will see system usage moving from small departments to company-wide and from basic applications to more business-critical ones.” Lopez’s consulting work is mainly with manufacturing clients, but he sees similarities in the life sciences.

The pattern of technology adoption noted by Lopez does seem to be prevalent in the life sciences when it comes to distributed computing systems. Most life science companies that deployed distributed computing systems have used them for basic applications, such as accelerating in silico sequencing calculations done in the basis research stage of a drug’s development.

Having seen the success of distributed systems in basic research, companies are looking to use their distributed systems for other areas of work, such as drug discovery and target validation. But to move into this new application area, many companies are looking for help from the distributed computing vendor community.

To that end, Platform, along with most distributed computing software vendors including AVAKI Corp., Blackstone Computing, Entropia Inc., and United Devices, will work with a company to port existing home-grown applications to the distributed environment.

But most vendors also offer versions of commonly used genomic research analysis programs optimized to run faster in a distributed environment.

For example, United Devices has optimized versions of HMMER Search, HMMER Pfam and BLAST that run on its MetaProcessor platform. Additionally, United Devices has partnered with Accelrys Inc. and offers optimized versions of Accelrys’ sequence analysis and target validation applications CHARMm, Delphi, LigandFit, Modeler, and SeqFold. And this year United Devices will add optimized versions of sequencing tools SIM4 and GeneWise.

Blackstone supports an optimized version of BLAST that runs on the company’s PowerCloud platform. Entropia supports accelerated versions of FRED and HMMER. Through a partnership with TurboGenomics Inc., Entropia supports TurboBLAST, an accelerated versions of BLAST.

The upshot for life science managers is that those who wish to use distributed computing as a company-wide resource will increasingly have the applications and management tools to do so.

 





For reprints and/or copyright permission, please contact  Jay Mulhern, (781) 972-1359, jmulhern@healthtech.com.