These tests gave researchers four objective measures of pigmentation to compare with their DNA blueprint analyses. Scientists then used a statistical model developed by Robert Valenzuela, a fourth-year genetics graduate student, to ascertain the three best SNPs to explain each pigmentation variation.
The resulting analyses enabled forensic scientists to determine hair, eye and skin color through a DNA sample. With DNA blueprinting, analysts found they can learn a person's hair color with 70-percent certainty. Examining three separate nucleotides can also pinpoint the ratio of brown-black to yellow-red in a person's hair with 43-percent certainty. Researchers also found analysis could explain skin color variations with 50-percent certainty, and that eye color could be predicted with 76-percent certainty.
"The bottom line from these studies is that when we look at a relatively small number of single-nucleotide changes, for each of these traits, we can make a fairly accurate prediction as to what a person looks like," Brilliant explains.
Putting DNA blueprinting to use
DNA found at a crime scene will first be analyzed and put into CODIS to see if there is a match. This analysis looks at 13 changes in DNA for identification purposes but does not examine SNPs. To learn a person's physical characteristics, technicians would have to perform a second analysis using DNA blueprinting.
"This does not in any way replace the CODIS system, which is very important in determining whether a DNA sample came from a specific individual," Brilliant says. "It wouldn't be used to convict an offender. Rather it would be probative evidence, in other words a lead that would help officers hone in on a suspect."
What does this mean to law enforcement? Plenty, says William Watson, chairman of the Association of Forensic DNA Analysis and Administrators.
With nothing more than a DNA sample, investigators can paint a picture of a person's physical characteristics. "Like all genetic markers we've developed so far, this will turn out to be forensically significant," Watson says, "It will be useful as long as it's used with the understanding that investigators should not limit their suspect pool to only those people fitting those characteristics."
Watson, a fellow with the American Academy of Forensic Science, cautions investigators to remember that genetics are only half of the equation with environmental factors making up the other half. He also reminds that eye, skin and hair color are all things that can be altered. Suspects can change their eye color by wearing colored contacts; alter their skin color through bleaching, dyeing or even tanning; and change hair color by dyeing it.
Rande Matteson, associate professor of criminal justice at Florida's St. Leo University and veteran law enforcement officer of 32 years, calls the current research promising, but questions whether it will be applied. "From a law enforcement perspective, we don't use the tools we currently have," says the retired FBI Academy instructor and crime scene technician. "If someone could figure out how to take all this wonderful information and use it in a way that catches bad guys, my hat's off to them."
Watson agrees, stating this technique may be limited, proving useful when police lack eyewitnesses or leads from state, local or national databases. "If you've got nothing but some genetic evidence, this method may be employed," he says. "If you've got other information, you may want to focus on more traditional investigative tools first."
The current DNA backlog presents another major concern. Through data collected from a large, representative sampling of law enforcement agencies in the United States, a National Institute of Justice-funded study in 2003 revealed:
- The number of unanalyzed DNA cases reported by state and local crime laboratories is more than 57,000.
- Total crime cases with possible biological evidence, either still in the possession of local law enforcement or backlogged at forensic laboratories, is more than 500,000.