• Politics
  • Diversity, equity and inclusion
  • Financial Decision Making
  • Telehealth
  • Patient Experience
  • Leadership
  • Point of Care Tools
  • Product Solutions
  • Management
  • Technology
  • Healthcare Transformation
  • Data + Technology
  • Safer Hospitals
  • Business
  • Providers in Practice
  • Mergers and Acquisitions
  • AI & Data Analytics
  • Cybersecurity
  • Interoperability & EHRs
  • Medical Devices
  • Pop Health Tech
  • Precision Medicine
  • Virtual Care
  • Health equity

A New Tool Uses DNA to Predict Eye, Hair, Skin Color


HIrisPlex-S could become a powerful force in healthcare, law enforcement, and beyond.

Image has been cropped and resized. Courtesy of Walsh lab in School of Science at IUPUI.

A snippet of DNA can now unearth a trifecta of information about a person’s appearance: the colors of their eyes, hair, and skin.

A forensics team at Indiana University-Purdue University Indianapolis (IUPUI), in collaboration with the Erasmus MC University Medical Center Rotterdam in the Netherlands, created a web tool called HIrisPlex-S that can predict the 3 hues using 41 genetic markers.

Even small DNA samples, such as those found at a crime scene or an archeological site, are enough to produce an accurate prediction, said Susan Walsh, PhD, an assistant forensics professor at IUPUI and co-director of the study, which was recently published in the journal Forensic Science International: Genetics.

>> READ: Personalizing Healthcare: The Netflix Approach

The team at IUPUI first created a tool called IrisPlex that predicted eye color in 2009, and 4 years later they added hair color. The newest version adds a categorical predictor for skin color divided into 5 hues: very pale, pale, intermediate, dark, and dark to black. The classification is based on the established Fitzpatrick Scale, which was created in 1975 to study the effect of ultraviolet rays on skin.

The HIrisPlex-S tool can help users generate a genotype with the relevant markers from a DNA sample, and it also provides a fill-in-the-blank tool where users can input the markers form an existing assay.

The predictive model was built on 41 pieces of DNA and trained on samples from a variety of countries and ethnicities.

Of the 41 markers, some contribute to one or more of the traits and have a range of effectiveness on the model. “If you're missing markers, that will impact on your prediction,” says Walsh, depending on the degree to which it is important to the model. “It will tell you if you can’t do it or not if you’re missing certain ones.”

Depending on the importance of the marker, it might make the model ineffective, or simply result in loss of accuracy, but a sample would need to be very degraded to make the tool ineffective, says Walsh. “There’s a ranking for each of the markers for the different traits. Some of them, especially for eye color, if you don’t have one of the major predictors then we won’t produce any prediction, because without that one, you really need it, so there’s no point in doing anything.”

The most obvious use case for the tool is when law enforcement can’t match a DNA sample to an individual—but HIrisPlex can be used to generate a description. It’s useful in both crime scene and missing persons scenarios, according to the experts

The tool has been used by law enforcement in the Netherlands, Poland, and Australia, but it has not yet been adopted in the United States, Walsh said. It has also been used on ancient DNA, and it was used to determine that King Richard III had blue eyes and blond hair.

Genotypes from commercial companies like 23andMe and Ancestry may have some overlapping markers, but not all have enough to input into the web tool.

The next step for Walsh is to move from categorical to quantitative prediction. “I’m trying to make it even more specific, so rather than say you have blue eyes, actually give the correct shade of blue,” she said.

Get the best insights in healthcare analytics directly to your inbox.


Why Consumers Could Drive the Genomic Revolution

The Democratization of Data Analytics

Genome Analysis Leads to New Insights into Opioid Addiction

Related Videos
Image: Ron Southwick, Chief Healthcare Executive
George Van Antwerp, MBA
Edmondo Robinson, MD
Craig Newman
Related Content
© 2024 MJH Life Sciences

All rights reserved.