Three Egyptian Mummies’ Faces Reconstructed Using DNA From 2,000 Years Ago
A group of scientists specializing in genetics in a laboratory in the United States has reconstructed in detail the faces of three Ancient Egyptian mummies from a community on the banks of the Nile from DNA sequences from more than 2,000 years ago. The authors of the analysis believe that this is the first time that advanced techniques for predicting an individual’s observable traits, phenotypic traits – in this case physical – have been used with such ancient human DNA.
The faces of the mummies, which correspond to those of three men, have been recreated with an appearance of about 25 years of age , and based on the assumption that their complexion was light brown, their eyes and hair dark and without freckles. In addition, the analysis of their DNA has also made it possible to predict that the three individuals had Jewish ancestry and roots from Yemen, Morocco and Tunisia respectively. The study has been presented by Parabon NanoLabs, specialized in DNA phenotyping services, at an international conference in this field held this month in Florida.
“We were all surprised and fascinated to see that DNA was no more similar to modern Egyptians,” explains Ellen McRae, director of bioinformatics at Parabon NanoLabs. “And that, of the three faces, only one looked Egyptian, the others were rather from southern Europe to me, and that is actually what we also see in the ancestry: that these people were, genetically, more similar to the people from the Mediterranean than people who are currently in Egypt ”, he slides.
The mummies from which DNA has been used came from an archaeological site along the Nile called Abusir El Meleq, located in the middle of Egypt and inhabited from at least 3250 BC to approximately 700 AD. The individuals in question lived at different times from the late New Kingdom to the Roman period of Ancient Egypt, and have been baptized with alphanumeric codes: JK2134, the oldest, dates from between 776 and 569 BC, JK2911 lived between the year 769 and 560 of the same era, and JK2888 is estimated to have done so around the years 97 and 2.
The results of the analysis, which imply that the ancestry of the three individuals was not sub-Saharan, are consistent with previous studies that had determined that ancient Egyptians shared more ancestry with the inhabitants of the Middle East than current Egyptians, since the latter received a additional sub-Saharan mixing in more recent times, according to the report.
“If you compare these individuals genetically with modern populations, their DNA was more similar to that of individuals from Yemen, Tunisia and Morocco, and not so much to that of people living in Egypt today,” says McRae. “They could have come from other parts of the Mediterranean, and they show no African ancestry, whereas modern Egyptians do,” he notes.
“If we can do this with DNA from 2,000 years ago, of course we can do it with one from 50 years ago.”
Raw data for mummies were obtained from European Nucleotide Archive (ENA), an open repository that provides free access to DNA data. From there, McRae explains that Parabon has a database of thousands of individuals with information on their DNA and what they look like. Thus, when they have a DNA sample from a person they do not know, as was the case with mummies, they use predictive models developed from the database to be able to determine their facial features. These were then compared to each other to discover differences, emphasize them, and combine them with pigmentation prediction so that a forensic artist could create the facial-looking compositions that have been presented.
The main challenge in this process, McRae notes, was that the DNA of individuals was very old and that, therefore, it was damaged by having been exposed in the environment for thousands of years, so that, even after removing the bacteria, a lot of sequence data was still missing. To guess the missing pieces of the puzzle, the group had a very advanced tool, called low coverage imputation, which allows them to statistically figure them out from the pieces around them that do exist. “As the data came in, we couldn’t have done this analysis,” says McRae.
“This really shows how powerful these techniques can be for difficult samples,” he says. “if we can do this with a DNA from 2,000 years ago, of course we can do it with one from 50 years ago, “he added.