There's also a question of whether, once a person starts smoking, increased nicotine consumption causes certain genes to turn on or off. To that end, Philibert plans to run similar genome-wide analyses on younger people who have not yet started smoking but may have a family history of nicotine dependence. He will also analyze DNA samples from patients with single psychological diagnoses to obtain what he calls a "cleaner phenotype."

Anthony Hinrichs, assistant professor of psychiatry at Washington University, in St. Louis, studies genes that code for certain nicotine receptors in the brain. He sees this latest approach as casting a wide net for other potentially related genes.

"Transcriptional profiling in general, and this study in particular, are very promising ways of nominating candidate genes," says Hinrichs. "I'm sure that these genes will now be on the map for other investigators."

However, Hinrichs believes it will be a while before scientists can design accurate genetic tests for such disorders. "So far, researchers have identified a number of genetic markers which may increase susceptibility to drug addiction," says Hinrichs. "Any of these would only increase the risk of addiction by a small percentage. We certainly have a long way to go before we can routinely use genetic tests."

Meanwhile, Gunter and Philibert recognize that this kind of genetic testing comes with a host of ethical questions.

"There are some people [to whom] you may be able to say on a genetic basis, 'You should never take the first drink,'" says Gunter. "And some will listen, and some may not. But it may have more impact on somebody if you are able to say, 'This is why you, as an individual, don't need to take the first drink: because you are at a particular risk for this to become a dependency, and this is how we know.' That's different from taking a profile and saying, 'This is another alcoholic; I'm not going to insure them.' Our part in this is to inform them of the level of their risk so it's truly an informed choice."