The algorithm is the subject of another study published in the journal Genome Research, in which U. geneticists, working in collaboration with a private firm, demonstrate the potential of software they call VAAST, for Variant Annotation, Analysis and Selection Tool.
"The average human genome differs from one another by about 3 million single-nucleotide variants. It becomes a needle in the haystack to determine which one of those is causing these diseases," said Mark Yandell, a U. associate professor of genetics and lead author on the second study. "Which one of the 3 million changes is it? That's where VAAST comes in."
An expected dramatic drop in DNA-sequencing costs could soon enable almost anyone to get their genome mapped, straining the medical establishment's ability to interpret these complex chemical sequences.
"VAAST can identify disease-causing mutations with greater accuracy, using far fewer individuals and more rapidly than was previously possible," said a news release quoting Lyon, now a geneticist and psychiatrist at the Children's Hospital of Philadelphia. "We are now applying VAAST to many other unknown conditions, including rare [inherited] disorders and other common disorders such as [attention deficit hyperactivity disorder] and autism."
Yandell and colleagues built VAAST in collaboration with Omicia, an Emeryville, Calif.-based software developer for which Yandell was previously a paid consultant. He still owns a stake of less than 5 percent in the company.
Not long ago, it cost millions to sequence a person's genome, or all of his or her genetic material. The cost is now about $10,000 and is expected to drop to $1,000 within three years, raising the possibility that genome sequences will soon be as common as MRI scans, according to Martin Reese, Omicia's CEO and scientific officer and a long-time collaborator with Yandell.
"We need to build the infrastructure around that and IT systems to interpret the data," he said. VAAST could lay some of the groundwork so that the massive public investment in genome sequencing will actually benefit individual patients.
"A genome has 3 billion base pairs. You want to narrow it to the 20 or 30 discrete variants that are relevant to you," Reese said. "Your doctor is interested in a certain disease, because you see him for a reason you have cancer or cardiovascular problems. What you want to know is what does the genome tell you about your cardiovascular profile?"
Reese and Yandell are co-inventors of the patented technology, and the U. is licensing Yandell's half of the intellectual property to Omicia to commercialize it. Reese plans a market launch in September, but he anticipates it will remain in a clinical research setting for at least a year.
"Because it's a probabilistic tool, with each human genome sequence it analyzes it becomes more powerful," Yandell said.
Added Reese: "We ultimately want to do predictive medicine. It's a little early for that. It will take time to understand which variant predisposes you to a certain disease. That's three to five years out."
Co-authors on Yandell's study include Reese and U. researchers Chad Huff, Hao Hu, Marc Singleton, Barry Moore, Jinchuan Xing and Lynn Jorde, all associated with the Eccles Institute of Human Genetics. It was funded by the National Human Genome Research Institute and the federal stimulus through a prestigious GO, or Grand Opportunity, grant intended to support transformative technologies.
Black did have healthy children, four girls and a boy. In an interview published on Nature.com, Black said she didn't dwell on why her sons died until one of her daughters gave birth to a boy who looked like her first son, born in 1979. Three of the daughters have since given birth to what the family calls "little old men."
"We didn't think that it passed on to the next generation. We didn't think that this would be a problem for them," Black said in the interview.