DNA ancestry testing increasingly used in law enforcement

Last year, I noted that British police used DNA to narrow down the racial ancestry of an unknown sex offender.

American police are not only using DNA-determined ancestry to narrow down suspects, but also to determine the racial ancestry of murder victims.

The child, dubbed "Precious Doe" by local residents, appeared to be black. But new DNA tests that can determine a person's heritage indicated she was of mixed ancestry — about 40% white. That meant she almost certainly had a white grandparent.
This year, a tip led police to an Oklahoma woman who had not reported her young daughter's disappearance. When the woman was found to have both a black and a white parent, police moved in. Further DNA tests determined that the woman, Michelle Johnson, was the girl's mother. Johnson and her husband, Harrell Johnson, the victim's stepfather, have been charged in the slaying...
In 2004, police in Charlottesville, Va., used ancestry testing to confirm the race of a suspect in six unsolved rapes that began in 1997. Police had been criticized for seeking DNA samples from local black men based on victims' descriptions of the assailant. The testing indicated that he indeed was of Sub-Saharan African descent.
Ancestry testing also has been used on a female skeleton that was found in the snow near Mammoth Lake, Calif., in May 2003. The slain woman initially was misidentified as southeast Asian, based on witnesses' descriptions of a woman seen in the area. DNAPrint Genomics found she actually was a Native American, a finding confirmed by analyses of her diet and bone composition and further DNA tests.

Looking back further, there was a Louisiana case in 2003. A serial killer's ancestry was apparently successfully profiled.

Setting aside the issue of just what is really being measured when racial proportion is claimed, overreliance on this technology may create new vulnerabilities. It's possible that juries may underestimate the potential for error:

But some defense lawyers say CSI and similar shows make jurors rely too heavily on scientific findings and unwilling to accept that those findings can be compromised by human or technical errors.

Errors could arise from causes as simple as samples being switched.