One man killed 12 people, raped 45 women and stole 120 houses in several towns in California (USA) between 1976 and 1986. For four decades, the crimes of the so-called East Zone Violator remained unsolved, until on April 24, 2018, when the FBI and the Sacramento Police announced they had arrested Joseph James DeAngelo, 72, in Citrus Heights, just outside the California capital.
After 40 years of failures, one of the researchers decided to introduce the genetic profile found in the crime scenes in GEDmatch, an open database on the internet where people put their DNA to find out their ethnic origins or find new relatives. The search yielded compatible coincidences with a common great-great-grandfather. Enough to rebuild the family tree and find DeAngelo.
A new research, published today in the magazine Science, suggests that it is already possible to identify half of the US population through these genealogical records of private companies. "The DNA of each person in these databases is like a beacon that illuminates hundreds of their relatives that are not in those databases," said geneticist Yaniv Erlich, the main author of the work and chief scientist of the group. MyHeritage, an Israeli company leader in the family tree business. At least 13 criminals have already been detained in the US with the same method as in the case of the East Zone Violator, according to Erlich's account.
"The DNA of each person in these databases is like a beacon that illuminates hundreds of their relatives who are not in those databases," explains Yaniv Erlich
The study has employed a database of 1.28 million people. Their results suggest that a genetic profile registry of 2% of the US adult population would be enough to locate the relatives of almost any citizen of the country, based on an anonymous DNA sample. With the current databases available on the Internet, used mostly by white citizens, it would be possible to identify 60% of Americans of European origin.
Erlich's group, which includes scientists from Columbia University and the Hebrew University of Jerusalem, shows that from the identification of a person with a great-great-grandfather, 850 candidates appear to be the culprit of a murder. If it is assumed that the crime was committed within a radius of 160 kilometers from the homicide's home, it is possible to exclude 57% of the family members. The approximate age of the suspect, calculated by witnesses, makes it possible to discard the remaining 91%. And knowing if it is a man or a woman leaves a list of 16 or 17 candidates, a figure that can be assumed for traditional research.
The authors urge the genealogy companies to establish measures to prevent this information from being used for spurious purposes. As a demonstration of the potential of this technique, the Erlich team has traced the DNA of an anonymous participant in a scientific project to their full name. "While politicians and citizens in general may be in favor of these forensic tools to solve crimes, we must bear in mind that they arise from databases that are open to everyone," the researchers reflect. "Therefore, the same technique could be exploited for harmful purposes, such as the identification of participants in scientific studies based on their genetic data," they warn.