SoylentNews is people
SoylentNews
The Orange County Register reports:
[...] one of California's most prolific serial killers and rapists was caught by using online genealogical sites to find a DNA match, prosecutors said Thursday. Investigators compared the DNA collected from a crime scene of the Golden State Killer to online genetic profiles and found a match: a relative of the man police have identified as [the suspect, who was arrested.]
[...] Authorities didn't give the name of the site, one of many, like Ancestry and 23andMe, that allow people to send in their DNA and find long-lost relatives. [...] Contacted Friday, representatives of both Ancestry and 23andMe.com said the sites weren't involved in the case.
takyon: Also at NYT, The Sacramento Bee, NPR, and CNN, which added:
When police announced they had finally caught the Golden State Killer, Bruce Harrington had a simple message for the politicians who fought his tireless efforts to expand the California's criminal offender DNA database. "You were wrong," he said.
Harrington, whose brother and sister-in-law were killed in 1980, spent years in front of public safety committees, pleading with them to embrace DNA technology. "And frankly I ran into a buzz saw of opposition."
Many state elected officials and rights groups fiercely opposed any attempt by the state to expand its DNA collection database. Critics cited the privacy rights of people in police custody and questioned the constitutionality of allowing the state to gather DNA samples without evidence of guilt.
In 2004, California voters passed Proposition 69, known as the "DNA Fingerprint, Unsolved Crime and Innocence Protection Act." It gave the state broader powers to collect DNA. Now, it could get samples from anyone not just convicted of a felony, but even arrested for one. In some cases, authorities could also collect DNA from misdemeanor arrests.
Say goodbye to your genetic privacy. We have killers to catch.
(Score: 3, Funny) by anotherblackhat on Friday April 27 2018, @04:51PM (4 children)
So they sent some random DNA to GEDmatch.com and got a "match".
If the chance of a mistake is a million to one, given there are 7 billion people on the planet, the odds are still 7000:1 in favor of this being a mistake.
I hope they found some other evidence that this is the right guy.
(Score: 1, Touché) by Anonymous Coward on Friday April 27 2018, @05:12PM
They did, according to the 4th through 6th paragraphs of the New York Times article.
(Score: 2) by frojack on Friday April 27 2018, @06:10PM (1 child)
ALL of their matches were "false positive" but they didn't find the criminal that way. They found some similar positives (well over a 100) that they then had to narrow down.
They then confirmed it by matching crime scene DNA to something this guy discarded, cup, can, Kleenex?
By that time they already had in focused in their radar.
But how many other people's trash did they dumpster dive into, and how many other crimes did they discover in the process. Parallel Construction Much?
No, you are mistaken. I've always had this sig.
(Score: 2) by anotherblackhat on Friday April 27 2018, @08:36PM
It still amounts to "we searched billions of people until we found a match"
Unless the probability of a random person with "matching" DNA is less than 7 billion to 1, the odds are still better that you've got the wrong guy.
It's possible that their test is that good (I've seen claims of trillions to 1 for DNA evidence) but I'm skeptical.
I hope that they can find something not DNA related to connect this guy to the crimes they're accusing him of.
(Score: 2) by Osamabobama on Friday April 27 2018, @06:54PM
Those made-up numbers are thought provoking, but the real numbers are too complex to summarize in a sentence. I will make up a process to demonstrate:
First, there were multiple samples of DNA evidence. Each one of those would have been tested to generate a profile. Those tests each had an error rate (that I will not attempt to guess). Those results were then correlated with each other with a confidence threshold, which is probably arbitrary. Then, that summary DNA profile was compared with the database at GEDmatch, where they have a small sample of the world's DNA. That match must then meet a threshold typical of extended family members (so not exactly arbitrary, this time). Once they have such a match, they identify an expanded pool of relatives that they can check via other means. That was about 100 men.
To use the standard
automotiveanalogy, they collected DNA, built aDodgeDart, then threw it at a wall full of dartboards, hit one, then started searching that haystack for the needle they wanted, found it, built another dart, and used that to nail the guy to the wall.Appended to the end of comments you post. Max: 120 chars.