Poodle DNA compared with that of mice, people.

Chihuahuas, Irish wolfhounds, pit bulls, beagles, greyhounds, and more. Man's best friend comes in a range of sizes, shapes, and temperaments unmatched by any other mammalian species. Biologists have now taken a step toward understanding that diversity by conducting a limited, but relatively quick and inexpensive, scan of one dog's full DNA sequence, or genome.

The data from this scan should ultimately help researchers study the more-than-300 human diseases, such as cancer and epilepsy, that also afflict dogs. The new work has already enabled scientists to compare the mouse, dog, and human genomes.

"The sequence of our genome is more similar to the dog's, despite the fact that the dog lineage split off first from the common ancestor," of all three mammals, says Ewen F. Kirkness of The Institute for Genomic Research (TIGR) in Rockville, Md., who led the dog-genome project. The rodent's unusually high mutation rate has made its DNA diverge more from people's than the dog's DNA has, he explains.

In the strategy pursued by Kirkness' team, biologists isolate copies of an animal's genome and break the strands of DNA into millions of short fragments. After determining the sequence of nucleotides making up each such piece of DNA, biologists use a computer to match overlapping sequences and piece together as much of the animal's full DNA sequence as possible. The more DNA analyzed, the better the chance that the final genome sequence will be accurate and have few gaps. For the human and mouse genomes, geneticists sequenced fragments equaling 6 to 10 times the DNA in the actual genome of each.

The National Institutes of Health in Bethesda, Md., is sponsoring a similarly thorough dog-genome project, but Kirkness and his colleagues wondered whether they could glean important information from a substantially smaller amount of DNA. If so, researchers might then consider sequencing the genomes of one animal from each of the 18 orders of mammals.

In the Sept. 26 Science, Kirkness and his colleagues describe their survey of the dog genome. They ultimately sequenced DNA equal to only 1.5 times the genome. From that work, they determined 77 percent of the dog genome and found canine DNA fragments corresponding to 18,473 of the 24,567 previously documented human genes. "We got more than we expected," says Kirkness.

The newly available dog genome is "just a wonderful resource," says Gustavo D. Aguirre of Cornell University.

The canine DNA analyzed came from a male standard poodle belonging to two of the coauthors on the new report, TIGR's Claire M. Fraiser and J. Craig Venter of The Center for Advancement of Genomics, also in Rockville. That selection isn't surprising given that Venter used his own DNA when his former company, Celera, performed its commercial sequencing of the human genome (SN: 5/23/98, p. 334). NIH'S doggenome project, scheduled to finish next year, uses DNA from a boxer.