U.S. flag

An official website of the United States government, Department of Justice.

Developing a Global Genotype Database to Predict Eye, Hair, and Skin Color from DNA

For the last decade, researchers have been steadily moving toward DNA phenotyping — predicting appearance by analyzing the unique features of their DNA. Two NIJ research grants have laid the groundwork for ongoing research into forensic phenotyping. 
Date Published
July 12, 2019

Forensic experts trying to give names to unidentified skeletal remains or solve cold case crimes are sometimes left with nothing more to guide their investigation than DNA evidence. If that DNA cannot be matched to an existing profile in CODIS — the FBI’s Combined DNA Index System — investigators are increasingly looking to predict the unknown person’s appearance using unique DNA sequences.

The most common features used for this analysis are single nucleotide polymorphisms (SNPs, pronounced “snips”), which are isolated mutations found across the genome. SNPs are easier to analyze than other mutations, and they can be correlated with variations in physical appearance. Forensic DNA phenotyping is also known as “predictive biometrics.”

Under her 2014 NIJ research grant, Indiana University biologist Susan Walsh aimed to correlate measurements of eye, hair, and skin pigmentation with new and known SNPs and whole genes. The project, she said, focused on areas “where improvement and fundamental research was required for more accurate prediction of pigmentation from DNA in both categorical and quantitative areas.” Such areas included understanding green and intermediate eye color, understanding age-related changes in hair color, and building systems for predicting categorical and quantitative aspects of skin color. 

The initial phase of Walsh’s work involved conducting a genome-wide association study (GWAS, called “g-wahz”) across 1,000 European genomes, looking for “the most predictive SNPs for quantitative eye, hair and skin color.” To increase the statistical outcome of the results, the researchers added more than 2,000 individuals from the United States. At one point, samples from Greek individuals were removed because sampling in that country proved difficult, and they were replaced by samples from 212 Lebanese subjects.

Finally, the scientists began work with a software company to develop a program “that enables users to input genotype (an individual’s entire genetic code) information to yield a highly specific quantitative color result for eye, hair and skin color.” They also began to work with collaborators to understand the genes involved in human hair structure (specifically whether it is curly, wavy, or straight), as well as predictive mutations for these traits.

Walsh noted that several jurisdictions currently conduct DNA phenotyping and the prediction of “externally visible characteristics” for law enforcement investigations. She believes that easy-to-use systems and tools would allow DNA phenotyping to become more widespread in law enforcement investigations. “With an increase in cold case investigations that have no success with standard [DNA] profiling, it is these new intelligence techniques that can really make or break a case in terms of new leads,” Walsh said.

Because of some of the work done under this project, she concluded, “eye, hair and skin color prediction is now available for forensic practitioners.”

Skin, eye, and hair color was also the focus of University of Cincinnati anthropologist Heather Norton’s 2013 NIJ grant. She proposed to expand our understanding of the genetic underpinnings of pigmentation by correlating it to SNPs in more diverse genomes than have typically been used, and to improve methods for describing pigmentation quantitatively. Norton’s team analyzed genomes from North and South American individuals who had African, Asian, and European ancestry, and they objectively measured pigmentation using a reflectance spectrophotometer for skin and hair. Eye color was taken from digital photographs and converted to a standard color coordinate system.

GWAS results from Norton’s research are still ongoing and suggest that the genetics of pigmentation in non-Europeans may be far more complex than previously considered. Also, although Norton’s team was successful in quantifying pigmentation traits, translating those measurements into broadly understood categories of skin, eye, and hair color is a further challenge.

About This Article

The research described in this article was funded by NIJ grants 2014-DN-BX-K031, awarded to Indiana University, and 2013-DN-BX-K011, awarded to the University of Cincinnati. This article is based on the grantee reports:

Date Published: July 12, 2019