|Dog Genome Points to New Sequencing Strategy|
By Kate Dalke
With a planet full of species waiting to be sequenced, a new study of the dog genome suggests a strategy for generating useful information about many organisms without going to the expense of completely sequencing each one.
The idea is to partially sequence an organism and then compare it to related species whose genomes have essentially been finished.
The notion of comparative genomics is not new, but the dog study illustrates what can be learned from a rough sketch of a genome. By taking the partial dog sequence and comparing it to the nearly complete human and mouse genomes, researchers learned about both dogs and humans.
They found, for instance, 18,000 genes in the dog that have human counterparts. There are about 350 genetic diseases common to both people and dogs, and often the same genes are involved in both species. In recent years research on human genetics has benefited dogs—and vise versa.
The three-species comparison revealed that humans are genetically more similar to dogs than to mice.
In addition, the study revealed thousands of short DNA sequences that are present in all three species but have no clear function. These mysterious sequences, sometimes derided as junk DNA, can now be isolated and studied.
“No one appreciated how much information you could get from this coverage of the genome,” says Ewen Kirkness of The Institute for Genomic Research (TIGR) in Rockville, Maryland, who led the sequencing. “There is a substantial amount of information to be gained.”
The new study was based on data generated two years ago, when the DNA of a male standard poodle, named Shadow, was sequenced at Celera Genomics in Rockville, Maryland—where the genomes of five humans and several mice had been sequenced.
Recently, Kirkness and his colleagues at TIGR and the Center for the Advancement of Genomics, also in Rockville, compared Shadow's sequence to humans and mice.
The study will be published in tomorrow's issue of the journal Science. Kirkness presented the findings yesterday at the 15th Genome Sequencing and Analysis Conference (GSAC) meeting in Savannah, Georgia.
The report comes at a time when genome researchers and funding agencies are weighing their options about new sequencing roads to take. Would it be better to fully sequence a few large—and therefore expensive—mammalian genomes, or sequence many smaller microbial genomes?
Shadow's genome was sequenced 1.5 times over. By comparison, the completely sequenced human genome has been sequenced about six times over. Kirkness estimates that four to six genomes could be “surveyed” for the cost of completely sequencing one.
The repetition is necessary because of the way genome sequencing works. In so-called shotgun sequencing, a genome is copied multiple times and then randomly shredded into small fragments that can be read by sequencing machines. The fragments are later assembled into an ordered genome by computers.
Because of the randomness of the sequencing process, the repetition is necessary to ensure that the entire genome has been sampled and properly ordered.
Using funding from the US National Institutes of Health, researchers are in the process of completely sequencing the dog genome. The DNA came from a boxer named Tasha who lives in upstate New York.
Some researchers have said they might not necessarily need the complete genome of their favorite species if “1x coverage” yielded enough data to make valuable comparisons to the fully sequenced genomes of other organisms. The information from Shadow's genome suggests that is the case.
For many organisms, a lower level of coverage will be just fine, says Elaine Ostrander, who studies dog genetics at the Fred Hutchison Cancer Institute in Seattle, Washington. For the last two years she and others have used Shadow's data in research.
The new report, Ostrander says, “nicely documents” how researchers can get the most out of a genome survey.
. . .