A Troubling New Direction for DNA Testing

What if DNA evidence from a crime scene could tell us what the perpetrator looked like?

Scientists around the world are working to tie our genes to the physical traits like eye color, height and race. Although this research has many important applications, one direction it has taken is the hunt for a composite sketch based on DNA. The inexactness of this science and the propensity of criminal justice agencies to use scientific methods before they're ready worries me. The U.K. and the Netherlands are already using some form of these tests, as are some U.S. states. Germany has outlawed the practice, along with Indiana, Wyoming and Rhode Island. But Louisiana is one state that has already used this testing, and it helped correct errors in an invesitgation. From the Wall Street Journal:

In 2004, police caught a Louisiana serial killer who eyewitnesses had suggested was white, but whose crime-scene DNA suggested -- correctly -- that he was black. Britain's forensic service uses a similar "ethnic inference" test to trace murderers and rapists.

I don't think it's a great idea to add another layer of evidence to the mix that's right only most of the time. But Slate's William Saletan wrote this week that we shouldn't be too quick to discount racial clues from DNA. If these tests are able to confirm or question other unreliable evidence, they should be included as a piece of the puzzle, he says.

Initial results of this method will probably be pretty crude. But will it end up being worse than old-fashioned police sketches based on eyewitness accounts? Would we have caught the Son of Sam killer earlier, for instance, if his police sketch hadn't been wildly inaccurate, making him look Latino or Asian?

No, DNA phenotyping isn't perfect. But it's better than nothing. And it's better than trusting witnesses alone.

The WSJ reports that researchers are able to predict eye color 70-90% of the time and skin color 46% of the time. These numbers aren't strong enough to rely on. Wrongful convictions happen - and real perpetrators get away - when faulty or limited science puts police on the wrong track in those crucial first few days after a crime. I agree that these tests could be helpful in some cases to confirm other evidence, but can we guarantee that genetic composite sketches won't be become simply another form of unreliable forensic science?