NEW YORK - The use of DNA from a genealogy database to identify a suspect in the Golden State killer case marks a turning point in forensics. Now, even if DNA from a crime scene doesn’t match any known suspect or anyone whose DNA was taken on a previous arrest, detectives have another recourse: They can search for a suspect’s relatives among the millions of people submitting DNA to trace their own genealogy. The number of such samples keeps growing each year, and the more data, the greater the odds of success.
In pursuing the 20-year-old Golden State case, detectives used DNA left by the infamous California serial killer to create a profile under a fake name on a genealogy site called GEDmatch.
That led them to some users who seemed to be distant relatives of the killer, and this data allowed a genetic genealogist to create a family tree. Ultimately, it pointed to a former police officer, 72-year-old Joseph James DeAngelo. He was arrested and is accused of over 50 rapes and 12 murders.
And again this month, genealogical DNA led to a new suspect in another case — the 1987 murder of a young Canadian couple traveling in Washington state. Other authorities are using the technique in similar unsolved cases.
Some are worried that using people’s DNA data this way threatens privacy, because these genealogical searchers can have unintended consequences not only for the people submitting the DNA but for their relatives as well.
The New York Times compared the use of DNA in the Golden State case to the way Cambridge Analytica surreptitiously used people’s Facebook data in an attempt to influence the 2016 election.
With crime solving, there are benefits to any technology that offers independent, verifiable lines of evidence.
Some experts say genealogical sleuthing could reduce the number of people imprisoned or executed for crimes they did not commit.
Greg Hampikian, a biology professor at Boise State University and founder of the Idaho Innocence Project, says he’s hoping genealogical data will help solve an Idaho murder case in which an early suspect, Chris Tapp, went to prison for 20 years despite the fact that his DNA didn’t match that left at the crime scene.
So far, police have not found anyone matching that DNA, which Hampikian said was left in abundance at the crime scene from a single source.
He said police tricked Tapp into confessing by promising him immunity if he would confirm the guilt of another suspect.
Scientists studying such cases say it’s surprisingly common for innocent people to confess when put under certain kinds of pressure.
Detectives ran into a similar situation in some of the very first cases in which DNA was used — the rapes and murders of two teenage girls in rural England in the 1980s.
As told in this historical account in the Guardian newspaper, police had become convinced that the perpetrator was a mentally challenged 17-year-old boy who had known one of the victims.
The DNA cleared the first suspect, and indicated the two young victims were killed by the same man, but left police with no clue as to who.
To solve the case, the police asked every adult man in the town to give a sample of DNA for comparison. Some people resisted at first, but concern about the at-large killer eventually led police to get DNA from 5,511 men. Initially no one matched, but eventually someone confessed he’d given a sample for someone else — a 27-year-old baker named Colin Pitchfork, who, once caught, confessed to the crimes.
That worked because the town was small enough, and the people willing to give their DNA up for the sake of stopping the killer. The growing DNA genealogy databases make this technique increasingly effective at a larger scale.
Hampikian, who has been an advocate for the wrongly convicted, said genealogy databases offer a source of important independent evidence. And yet, he’s also been vocal in pointing out the weakness of DNA evidence when scientists have only tiny traces left when people touch an object.
In those cases, he has said, it’s easy for investigators to contaminate objects, as he believes happened in the case of Amanda Knox, an American exchange student who was convicted and then exonerated in the murder of a young woman in Italy.
But if there’s enough DNA for a standard analysis, the odds of a false match are minuscule. As for the privacy concerns, he suggests people who give up data and allow its use in investigations become eligible for reward money.
At the very least, those collecting DNA for genealogy might give people the chance to opt in as a public service — a way to prevent real killers from taking more victims and to keep innocent people out of prison.
Science writer Faye Flam is a Bloomberg Opinion columnist.