Computers Tackle Sea's Ecology

PETERBOROUGH, N.H. — A new approach to the study of the sea's ecology will use high-speed gene-sequencing techniques combined with data mining and optical networking in an attempt to give biologists worldwide a window into the life of the oceans.

Biologists and computer scientists are meeting this week at San Diego's Birch Aquarium to launch the joint project involving the Scripps Institution of Oceanography, the J. Craig Venter Institute and the Calit2 computing facility at the University of California, San Diego.

"In the ocean, you have a very dense ecology of microbes, particularly in the surface," said project leader Larry Smarr, the director of Calit2. In the initial phase of the project, samples of ocean surface water will be taken, the microorganism found there will be broken down, and their DNA will be extracted and sequenced.

"It's a shotgun approach," Smarr said. "You put the microbes through a blender, and then sequence the fragments of DNA that are floating around. Then you use a lot of computing power to reconstruct the genomes of this collection of organisms."

The Calit2 facility has been building a new type of 10-Gbit/second optical network designed to tie supercomputers, terabyte data storage servers and wall-sized high-definition displays together. The facility is also linked to the National Lambda Rail optical-computing network, as well as to similar scientific computing networks in Europe and Asia. Later this year, the data from the ocean survey will be available via the Lambda Rail network, so that biologists will be able to analyze and visualize the genetic ocean data remotely. Lambda Rail is a dense wavelength-division multiplexed optical network that is being financed and installed by computing centers across the country. Biologist J. Craig Venter pioneered the "shotgun" approach to sequencing genomes in the race to sequence the human genome. In Venter's method, sequenced islands of the genome are first created and then fitted together later into the entire sequence. The nonprofit J. Craig Venter Institute, which Venter formed to apply his methods to solving problems in biology, last year began a program to catalogue the organisms in the air over cities. That project filtered microorganisms from the air atop office buildings in New York to sequence their DNA and set up a reference database that may help the medical community track pathogens.

"What he [Venter] realized is that the DNA in microorganisms is about 1,000th the length of the human genome," Smarr said. "Instead of 3 billion base pairs, it is more like a couple of million-so it is a thousand-times-simpler problem."

That means it would take the same amount of work to sequence 1,000 microbe genomes as it took to sequence the human genome, which is now a routine operation. "They literally throw a bucket overboard and bring it back up and take that water, with probably several thousand species of microbes, and they shotgun-sequence it and reconstruct what all the species were," Smarr said.

Most of what is known about the world of microbes has been derived from soil studies, which have led to medical revolutions, such as antibiotics. Smarr expects similar discoveries will be made from the joint project's survey of the ocean, discovering medically or industrially useful organisms.