Coriell is also building a Web portal, due to launch in September, where participants can access their genetic-risk information and elect to share some or all of it with their doctors. They will also fill out regular health surveys to allow scientists to assess how useful this genetic information is in improving long-term health: lack of information on clinical utility has been a major criticism of direct-to-consumer genetic screening. The portal will also feature educational material targeted to physicians, including continuing-medical-education courses.

When the site launches, will physicians know what to do with the genetic information their patients provide them with? "Absolutely not," says Christman. "But it's where we have to start. We can't wait to train a new generation of doctors."

Preliminary surveys that Coriell has conducted suggest that patients are also confused by genetics. "I'm alarmed by the lack of understanding," says Christman. "People don't understand the difference between Mendelian and complex genetic diseases." In a Mendelian disease--named for the 19th-century priest Gregor Mendel, who did pioneering genetic research on fruit flies--the disease is linked to a single gene. Carrying certain variants of that gene virtually guarantees the eventual onset of the disease. With complex diseases, however, multiple genes interact with the environment to increase risk. Carrying any particular gene does not guarantee disease onset. "I don't want people going home and saying, 'I have the heart-attack gene,'" says Christman.

Unlike some private services, the Coriell project will give participants only that genetic information deemed by an advisory board to be medically actionable, meaning that there are medical or lifestyle interventions that can reduce risk of the disease or improve outcomes. (Of course, the definition of "medically actionable" can vary. Navigenics also limits testing to disorders for which interventions exist. But it includes Alzheimer's disease on its list. Christman says that Coriell does not, since no interventions have been proved to have a significant impact on Alzheimer's risk.) An independent review board will meet twice a year to discuss the latest research and decide which variants make the list.

As part of their efforts to engage the medical community, Christman and his team have reached out to local hospitals, which are offering the genetic-screening service to their staff, both as an individual benefit and as a way to learn about genomics. "The project is extremely important because it will add a dimension to patient care we haven't had before--the genetic component," says Simon Samaha, chief medical officer at Cooper University Hospital, in Camden, one of the participating organizations. Coriell has enrolled 2,000 participants to date, about half of them medical professionals, and it aims to have 10,000 volunteers by 2009. Currently, only people who live locally, and thus can give consent to participate in the project in person, are eligible.

Because participating is free, Christman says, the project will reach a broader cross section of the population than do private companies, which charge $1,000 to $2,500 for their services. Scientists at Coriell have directly reached out to both African-American and Hispanic communities.