May 9 BusinessWeek article. DoubleTwist softwares found 1,500 genes vs. 545 on Chrm 22. I think DT's softwares are either super smart, or they are just printing garbage. If those 545 genes are all in their 1,500 findings, then they can help to limit the search.
Has DoubleTwist Unraveled the Genome?
The startup says its powerful software can locate the small number of important genes in the great mass of "junk DNA"
NEWS ANALYSIS
MAY 9, 2000
businessweek.com
A dark horse has emerged in the race to decipher the human genome -- the genetic bits that determine eye color, body weight, and a host of other personal traits, including disease susceptibility. DoubleTwist, an Oakland (Calif.) Internet startup, announced on May 8 that it has used a series of powerful computer programs to generate a rough map of the probable locations of the 100,000 genes contained in the genome. The company plans to make the information available for a fee at its Web site, www.doubletwist.com.
So far, this has been largely a two-horse race. Since January, the Human Genome Project, a network of government-funded labs, has been running neck and neck with Celera Genomics Corp., the Rockville (Md.)-based biotech company run by gene wizard J. Craig Venter, to unlock the human genetic code. Any day now, Celera is expected to announce that it has completely sequenced the entire human genome (see BW Online, 4/7/00, "Craig Venter: The Bad Boy of Genomics Makes Good").
And on May 8, the public project announced the official launch of the final phase of the human genome sequencing. The project will generate a "working draft" of the human DNA before producing a "finished" sequence.
JUNK SORTING. But sequencing the genome will be just the start. Linking raw genetic information to specific variants that cause a multitude of diseases, such as diabetes and Alzheimer's, will take years more of research. The next task is to cull through the more than 3 billion bits of DNA sequence to identify the 3% to 5% of DNA that actually play a role in producing such variants. (Scientists refer to the rest of it as "junk DNA.") It's here, though, that DoubleTwist may have a head start on the competition.
Celera is already hard at work developing a suite of powerful computer tools that can interpret the vast amount of gene-sequence data the company is producing. It's just one of roughly a dozen public or private centers using high-powered software to identify potential genes. But only DoubleTwist has announced the completion of a truly comprehensive first draft, putting increased pressure on Venter and the others to continue to update and develop their own tools.
DoubleTwist was formed last December as an Internet portal for medical and biological researchers. At the company's site, scientists can analyze DNA sequences, monitor their patent status, and design experiments to test the roles of newly discovered genes. Recently, the company teamed up with Sun Microsystems to build a powerful computer network capable of differentiating genes from junk DNA in the raw data that has already been made public, roughly 85% of the total sequence.
TRIPLE PROOF? According to Nick Tsinoremas, research director at DoubleTwist, the company has identified the locations of some 65,000 genes with about 90% accuracy. His researchers have used three different sets of software to analyze these particular genes -- and all of them have produced the same answers. The company has also detected an additional 40,000 genes, but the data associated with this second set are more questionable -- Tsinoremas believes they're only 50 to 75% accurate.
To demonstrate the power of their software, Tsinoremas and his colleagues chose chromosome 22 as a test case. The sequence of chromosome 22, which is the smallest and most gene-rich human chromosome, was published last December in the journal Science. At that time, researchers reported identifying 545 potential genes. Starting with the same raw sequence data, the DoubleTwist scientists claim to have discovered nearly 1,500 genes with their software. The results will be formally presented at a scientific meeting in Cold Spring Harbor, N.Y., later this week.
Because the research hasn't yet been published in a scientific journal, experts are cautious in their support. "I know they have some good people there, and they've probably done a good job of it, but I can't say more until I've seen the data," says David B. Lipman, director of the National Center for Biotechnology Information in Maryland.
"OLDEST BUSINESS." DoubleTwist plans to make its gene list available to researchers for a yearly subscription of $10,000. CEO Couch says that while this might seem expensive, it's far cheaper than the deals that Celera and Incyte, another genomics company, have already sealed. DoubleTwist also plans to offer a discount subscription to university researchers.
It's not clear yet whether DoubleTwist can survive as a business selling its information so cheaply. William Haseltine, CEO of Human Genome Sciences, a genomics company with three drugs in clinical trials, fears DoubleTwist's business model because "the margins are too low in that kind of commodity-based business." That's why his company's mandate of his company has always been to sell gene-based medicines. "It's the oldest business in the world," says Haseltine. "What we are doing is using the body's own substances to regenerate what is damaged by disease, injured by truama, and eventually worn by age."
A shakeout of biotechs and genomics players may lie down the road. But regardless of which survive, the knowledge these companies are generating in mapping the genome is sure to lead to better medicines.
By Ellen Licking in New York
EDITED BY DOUGLAS HARBRECHT |
