DNA evidence was first used to help solve crimes in 1986, but it didn’t hit the main stage of the public’s attention until it was used in OJ Simpson’s murder trial in 1994. This relied on DNA fingerprinting, which is now commonly used in trials. One unfortunate flaw is that it requires a suspect - without a person to match it to, DNA evidence doesn’t mean a whole lot.
The field of genetics has come a long way over the last few decades, so now could we actually use DNA found at a crime scene to locate the suspect, even without any leads? A new study led by Mark Shriver of Penn State has shown that rough 3D mugshots can be built from genetic samples by using a mere 24 genetic variants. The results have been published in the open access journal PLOS Genetics.
Human facial features rarely rely on just one allele. There are approximately 150 genes that affect skin, eye, and hair tones. Expression of the genes is influenced by several genetic and environmental factors including climate. However, a relatively small number of genetic variants can create a close approximation of a person’s face.
Shriver’s team collected genetic samples and high resolution photographs of 592 volunteers living in either the United States, Cape Verde, or Brazil who were of European or West African descent. After making a mesh from over 7,000 landmarks on the face, the team was able to compare shapes of facial features against one another.
Image credit: Shriver et al.
After the faces had been categorized, the team focused on identifying the genes that are involved with the embryonic development of those features. They identified the single nucleotide polymorphism (SNP) that varied in these genes. In all, they found 24 SNPs that spanned 20 genes which were most likely to determine face shape and features.
This isn't the first time that facial reconstruction has been attempted from genetic samples. In 2012, a group of researchers from the Netherlands analyzed DNA from over 10,000 Europeans and found SNPs that connected nine facial features, including eye width and the shape of cheekbones. Additionally, Kun Tang from the Shanghai Institutes for Biological Sciences has been able to identify 5 SNPs related to facial features which are unique from the ones discovered by Shriver’s team. This research involved a scan that used over 30,000 data points per face, giving a much more detailed representation of the features.
In the future, Shriver is hoping to identify SNPs associated with hair texture and those influenced by gender. More volunteers of different ethnicities will be required to identify additional SNPs involved in the development of facial features. It is hoped that this technology could be used to help give police a better starting point when seeking suspects. Additionally, it could help identify unknown victims or be used by archeologists to get a better idea of what ancient people looked like throughout history.
Image credit: Shriver et al.