By Kim Coghill

Washington Editor

WASHINGTON - Celera Genomics Group and Sandia National Laboratories Friday said they entered a multimillion-dollar four-year agreement to develop the next generation of software and computer hardware for computational biology and life sciences applications.

Representatives from both organizations, as well as Houston-based Compaq Computer Corp., which will provide the technology, formally signed the agreement Friday at the Department of Energy with Secretary Bill Richardson presiding. Sandia, of Albuquerque, N.M., is a multiprogram DOE laboratory that has major research and development responsibilities in national security, energy and environmental technologies.

"This is a development program designing future computers to work around holistic biology use, so the initial phases are actually computer design and algorithm development that scientists from both Celera and Sandia will be working on developing," Craig Venter, chief science officer and president of Rockville, Md.-based Celera, said during a conference call with reporters and investors. "Algorithms not only improve the speed of the project, but take it to the next phase of development."

The venture is expected to result in computer programs that can better decipher information, including the complete map of the human genome.

Venter would not elaborate on the cost of the project except to say, "The total contribution on all sides is multimillion dollars."

Celera's stock (NASDAQ:CRA) closed Friday at $41.062, down 6.25 cents. Compaq's stock (NASDAQ:CPQ) closed at $20.50, up $1.31.

The companies will work toward increasing computing capability with the goal of achieving 100 trillion operations per second (100 TeraOPS). By sharing some computing technology developed by Sandia, Celera and Compaq may reach the "petacruncher" (1,000 TeraOPS) level.

Celera and Sandia will concentrate on creation of advanced algorithms for biology research and on new visualization technologies for analyzing the massive quantities of experimental data from high-throughput instruments.

The driving design for this next-generation supercomputer are anticipated computational and data management requirements for proteomics. These requirements are expected to be more complicated than the pattern recognition and assembly operations required to sequence the human genome, according to a statement released by the companies.

"I think it is important to stress how we came to develop this cooperative research and development agreement," Richardson said during the conference call. "We were each running down similar paths to bigger and bigger supercomputers, and we were each working on a problem that is much bigger than we can take on ourselves."

Richardson said the collaboration provides both Sandia and Celera the opportunity to exchange expertise and, "in my mind, that is much better than just being funded to do research, because as advances are made on one side, they get applied by both organizations and the pace of development and research is superior."

Venter said Celera and Sandia will retain individual rights to patented materials brought to the collaboration, but will jointly own products that result from it.

Venter, along with National Institutes of Health Director Francis Collins, last summer flanked former President Bill Clinton when he gave both scientists credit for unraveling the genome. But Venter often is credited with success of the project because of his use of sophisticated computer technology to speed up the process.

Although both Venter and Collins previously have denied competing to produce the human genome map first, popular culture suggests otherwise. Work on the publicly funded Human Genome Project began 10 years ago, while Celera, a three-year-old company, started working on the project in late 1999. (See BioWorld Today, June 27, 2000.)

Celera's goal is to become the definitive source of genomic and related medical information. The company also is sequencing the mouse genome.