IF YOU'RE CONVICTED of a felony (or in some states a misdemeanor), your DNA goes into a database. That information primarily helps in the pursuit of repeat offenders. But some people want to extend the reach of that data to find people who are only a partial match. It's a particularly personal form of a law enforcement fishing expedition.
The technique is called "familial searching," and it targets not only the convicted but their relatives as well.
Sometimes, when an investigator tries to match a crime scene sample to the several million profiles in, say, the FBI's database, no exact match turns up. But there might be someone whose DNA profile is unusually similar. If the partial match is sufficiently close, or if some of the genetic markers in the sample are sufficiently rare, it could mean that the crime scene sample was left by a close genetic relative of the person who is included in the DNA database. Thus the familial search casts suspicion not on the convicted criminal in the database but on that person's siblings, parents or children.
Should forensic scientists reveal partial matches to police and prosecutors? Should officials be able to use this DNA as the basis for investigating relatives? Is this a lead that any investigator would be crazy to ignore, or is it an encroachment on civil liberties?
The difficulty is that it is both. While mining the DNA database for clues is certainly tempting, it is a temptation we should resist. Fairness and privacy concerns require it.
Although Britain has been using familial searches for several years, it's just arriving in the United States. Until recently, the FBI refused to allow states to reveal partial matches in its database to other investigatory agencies, but it's beginning to soften that stance. Many prosecutors, including those in California, are lobbying hard to be able to use the technique, and a 2006 article in Science argued that the use of such kinship analysis could increase the number of cases that were solved by up to 40%.
It's easy to see the arguments for using familial search techniques. If we forgo partial matches, violent criminals may remain at large. There are some powerful anecdotal examples of how familial searching has led police to suspects. And we use partial information all the time in other settings. If someone looks at suspects in a photo spread, for example, and says, "It's not any of those people, but the perpetrator looked a lot like No. 3," any competent investigator would think to ask if No. 3 had a brother.
