.Back to listing

Who Is It, Whodunit?


Less than a year ago, a New York artist gained unexpected publicity for using genetic profiles of strangers’ discarded trash to reconstruct facial features and create sculptures. Provocative art (perhaps) somewhat creepy (yes), and possibly illegal (in some states). But also, at the time, the predictive capacity of a few genetic markers for developing recognizable portraits was fairly weak.

Now scientists at Penn State U and in Belgium have taken the genetics to a new level, using superimposed stereoscopic camera images to pinpoint the precise location of facial points based on genetic markers and developed a statistical model that combines these genetically determined measures and includes ancestry informative markers (AIMs), sex, and race. The model builds on Manfred Kayser’s research on biological bases of externally visible characteristics (EVCs). His team has already developed several EVC markers to predict eye color, hair color, and five genetic variants to determine facial shape.

In criminal investigations, forensic efforts to match DNA profiles to known reference samples from suspects or those in a databank have reached a utility plateau with limited return on investment in expanding DNA databanks. Law enforcement needs more creative ways to solve crimes than expanding a databank and this investigative power of genetic information is another step closer to improving science over traditional witnesses.

I, for one, am excited to see this science mature and anxiously await how the law enforcement community integrates this into practice. I am intrigued to learn the societal effects of this technology, particularly the repercussions of surreptitious collection of public citizens to confirm genetic identity of a suspect to an evidence item.

But I am most excited to see this technology used in mass disaster victim identification. Whether caused by a terrorist, war, or an act of nature, thousands of people die each year and will remain nameless remains without science for identification. But dental technology and DNA are only as good as the reference information provided. The ability to predict aspects of identity without a reference sample is foundational for humanitarian identification efforts.