The banks of genealogical data are in the process of establishing itself as a tremendous tool for police services, but their effectiveness is such that they begin to pose serious dangers to privacy, according to a study published Thursday in the scholarly journal “Science”.
With online data banks such as GEDmatch, the results of genetic tests sold directly to individuals make it possible to trace cousins in the third degree (or closer) in 60 % of cases, and the crossover with other information, such as age or gender leads to results surprisingly sharp, which could be misused.
Thanks to spectacular advances in technology to “read” genomes made over the past 20 years, several companies such as AncestryDNA and 23andMe are now offering genetic tests to consumers wanting to know where were their ancestors. To this are added to Web sites like GEDmatch or DNA.Land, which allow you to upload the results of these genetic tests for further analyses.
In recent months, police departments have begun to use these sites for tracking down wanted criminals for a long time, with sometimes spectacular results (see two examples below). It is for this reason that the researcher Yaniv Erlich, affiliated with Columbia university and of Jerusalem, and three of his colleagues to test how far these new tools allow you to go.
Essentially, they are a part of a sample of 1.28 million genetic tests. They have knowingly removed the pairs which were similar in the older (siblings and cousin in the first degree) to take account of the fact that these tests often do in the family. But despite everything, they managed to identify one or more cousins to the second or third grade in 60%.
And by crossing these information with the geographic data (region where it is suspected a person of living) and demographic characteristics (approximate age, and sex), they are able to identify individuals. They were also able to identify a woman who had participated in the scientific project 1000Genomes and in which the genetic information is not public.
“Although policy makers and the public can see a good eye the fact that these new capabilities medico-legal help solve crimes, it remains that they rely on data banks and online services that are open to the public. Thus, the same techniques could also be used for purposes harmful, such as re-identification of people who have participated in research projects from their genetic data,” warns the article of Science. This does of course not prevent the police from doing its job, but rather of the restrictions to limit abuse.
TWELVE MURDERS, 50 RAPES, 40 YEARS Of IMPUNITY
The investigator Paul Holes has tracked down the Golden State Killer for nearly 25 years now, without success. For a dozen years, from 1974 to 1986, this serial killer had raped over fifty women and murdered 12 people in several regions of California. His DNA was found on the scenes of several of these murders and others being awarded by modus operandi) and on rape victims.
But as the culprit had never been filed by the police before they commit these crimes, the genetic information was not used to much. Mr. Holes has followed hundreds of tracks over time, which led to all cul-de-sac. Last year, however, he compared old samples of DNA of the killer to that found in banks of genealogical data. He had nothing to lose by trying, after all.
Mr. Holes has created a fake account on the site, GEDmatch, has entered the DNA of the murderer, and the results were not long : without lead directly to the suspect, the exercise has given him a list of at most twenty distant cousins. This was already much better than hundreds of potential suspects. Going back the generations, the investigator and his team have identified the common ancestors of all of this lineage — a couple of the early Nineteenth century.
And then they have laboriously recreated a family tree of their descendants. This accounted for about 1,000 people, but given what they knew of the Golden State Killer (a man, had to be at least towards the end of adolescence at the beginning of the crime in 1974, had to have a certain relationship with the Sacramento area in the 1970s), there were more than two.
One was eliminated by the DNA of a close relative. The police took the other by spinning and then, when he magasinait, were able to take a DNA sample on the door handle of his car. It was the same as the one linked to the murders and rapes. Joseph Deangelo, age 72 years, a former police officer, was arrested in April. His trial is ongoing.
Sources : NY Times, USA Today, Washington Post
ALL THIS IS WHAT WE FIND IN THE GARBAGE…
It was a long time that we knew the DNA of the killer of little April Tinsley, raped and murdered at Fort Wayne (Indiana) in 1988 at the age of eight years.
His semen was found on the underwear of the little girl. Particularly devious, the man had also left threatening letters to other small in the same region during the following years, saying they were “next” on his list; these letters often came in plastic bags containing condoms worn-out whose DNA was the same as that found on the body of April Tinsley.
But here, the DNA can be used to identify a suspect that if he is already in the banks of genetic information to the police (ex-prisoner, offender monitoring, etc), and this was not the case of this man. For 30 years, the investigation yielded nothing.
Until the police in Fort Wayne to use the services of a new company, Parabon NanoLabs, which uses banks of genealogical data to identify suspects.
The killer in question was not contained therein, but he had relatives who had already passed genetic testing and who had uploaded online. This would reduce the already enormous list of potential culprits and, by cross-referencing this information with others, the police was able to focus its efforts on two brothers living in the region of Fort Wayne. One of them lived in a trailer park.
Investigators have searched his trash, they found three condoms are worn, in which the DNA left no doubt : it was the culprit. The american media announced his arrest in last July.
Source : Washington Post