DNA used to recreate 3D face images in groundbreaking new investigative technique

A new technique could use a person’s DNA to reconstruct a 3D model of their face.

Researchers have devised a computational tool, called Difface, that uses the differences between single letters of the genetic alphabet (SNPs) to create so-called “3D facial point clouds.” These are sets of data points that represent the exterior of a face.

The results are published in the journal Advanced Science and the new technique could potentially be used for crime solving as it essentially allows for the recreation of an unseen suspect’s face image through DNA.

The researchers tested the technique on a Chinese database that held single nucleotide polymorphisms (SNPs) and showed that Difface could create an accurate facial reconstruction using DNA data. When information such as age, sex, and body mass index was added to the mix Difface improved the accuracy of the 3D images.

“Amazingly, Difface could generate 3D facial images of individuals solely from their DNA data, projecting their appearance at various future ages,” said Luonan Chen, an author of the paper and a researcher with the Chinese Academy of Sciences.

While the researchers claim Difface shows “excellent performance” in DNA to 3D image reconstruction, the method is not yet complete. “Validating Difface with datasets from multiple ethnic groups and exploring whether additional genetic loci are necessary for certain facial features will be key steps to ensuring that the model generalizes effectively across diverse populations,” according to the paper.

In other words, it might not work as well with people from different ethnicities. The biggest issue using the technique however is ethical as the paper observes. “By enabling the prediction of physical traits from genetic information, DNA phenotyping intensifies concerns about the misuse of sensitive personal data, which could lead to genetic-based discrimination and privacy infringements,” the paper claims.

Forensic DNA phenotyping (FDP) has emerged as a groundbreaking tool in criminal investigations, generating leads in cases where traditional methods have stalled. But the application of FDP raises many concerns, critics say. While FDP has made strides in predicting traits like eye and hair color with reasonable accuracy, predicting facial morphology is shakier ground and risks wrongful suspicion of individuals and groups of people.

In 2022, Canada’s Edmonton Police Service asked the public to help identify a suspect from a computer-generated image based on DNA phenotyping, but quickly walked back the request. The suspect was sought for an unsolved sexual assault in 2019 in the Canadian city. The image was generated by Parabon NanoLabs, which has received criticism related to lack of transparency over its methods.

Article Topics

biometric identification  |  biometrics  |  dna  |  face biometrics  |  forensics