The at-home DNA test craze is putting us all at risk

[jedishrfu]

Interesting video on DNA and privacy in today's world:



Specifically folks in some instances can be identified from their relatives DNA.

The worry of course is false positives placing you at a crime scene or of a sketchy insurance company profiling you and refusing coverage based on what it pieces together in your DNA makeup.

Some states are proposing restricting the use of familial searches by law enforcement because of privacy concerns.

 

[DrClaude]

Considering that these tests are basically worthless to learn about your ancestry, it is definitely not worth the risk.

 

[russ_watters]

Considering that these tests are basically worthless to learn about your ancestry, it is definitely not worth the risk.

People do a lot of "basically worthless" things without considering risk much. Moreover, "risk" tends to be a handwavey thing. More on that in next post.

My main concern would be the misleading advertising about "accuracy". Oddly, the video you linked didn't really cover well except for a vague mention in the last minute. The primary issue is that "genetic ethnicity" isn't a precise thing, so it should be obvious that it isn't possible to measure it precisely. The methodology uses a sampling of a bunch of people from a certain region and measures the frequency of certain genes. The problem with this is obvious: people move. The populations aren't static. This is most obviously expressed in the fact that "Ethnic American" isn't a thing in these results; we're too diverse and everyone is assumed to have heritage somewhere else. But what; don't people in Europe move? Maybe 1000 years ago, populations were more static, but they were never completely static. I'm part Scottish --- or am I? I have a fairly recent ancestor from the Orkney Islands, on the north end of the island chain. This area was alternately settled and annexed back and forth by Norway and Scotland. So am I Scottish or Norwegian?

It's important to note, though, that there are actually two separate purposes for these tests. One is the genetic ethnicity, but the other is for finding close relatives and building your family tree.

[edit] And now Facebook thinks I should vacation in Scotland...

 

[russ_watters]

Specifically folks in some instances can be identified from their relatives DNA.

The worry of course is false positives placing you at a crime scene...

As I said above, there's "risk" in everything, and the word is often thrown around in a handwavey way -- the title of the video is pretty provocative and while they identify the risks, they make no attempt to quantify them or analyze them in context vs other similar risks.

I'd be curious to know if a false-positive DNA match has ever happened. I found one link that suggests a certain test that is pretty prone to false-positives (1 in 700something) once got a guy wrongfully convicted in Taiwan, but that's all I can find. A more common test, though, has a theoretical false positive rate of about 1 in a billion. That's lower than the odds of winning the Powerball and I would hope some basic police work would prevent most such false positives from going anywhere. E.G., I'd hope they wouldn't arrest me for a murder in 1981 in California given that I was a 5 year old living in Pennsylvania at the time.

Is this really an actual worry people have?

... familial searches...

This is real, but I wonder what the actual "worry" is, and I'm not sure if it is calculable. If my brother (I don't have a brother that I'm aware of) kills someone and my dad's DNA pops up as a familial match, police will investigate me and my brother. This could include knocking on my door and questioning me. This is new, but is not any different from classic police work. If someone steals my car and uses it to commit a crime, the police will question me about it. If I'm at a restaurant and someone steals money from the cash register but they aren't sure who, I'd expect the police to track me down from my credit card slip and question me. Police (detective) work is all about making connections. The dna family search is just another entry-point. If our culture has people so afraid of the police that they are afraid of being one of the dead-ends in an investigative thread, that's a problem that doesn't have anything to do with DNA.

...or of a sketchy insurance company profiling you and refusing coverage based on what it pieces together in your DNA makeup.

This "risk" is entirely hypothetical, but paradoxically probably the bigger "worry". The odds of this happening tomorrow(without recourse) are exactly zero since it is currently not legal to do this. The real "worry" I think should be a law being passed that makes it legal. It would be tough to calculate the odds of this ever happening, but I suppose if people want to worry about it, they're entitled to.

I'm not much of a worrier.

 

[Ygggdrasil]

The tests may also be worthless at providing useful medical information as well:

A study of 100,000 people released earlier this month suggested that this experience could be widespread. Nearly 90 percent of participants who carried a BRCA mutation would have been missed by 23andMe’s test, geneticists found.

23andMe’s testing formula for this risk is built around just three genetic variants, most prevalent among Ashkenazi Jews. The new study demonstrated that most people carry other mutations of the gene, something many doctors have long suspected.

“It’s as if you offered a pregnancy test, but only the Jewish women would turn positive,” said Dr. Munster, who is the co-leader of the Center for BRCA Research at the University of California, San Francisco.

https://www.nytimes.com/2019/04/16/health/23andme-brca-gene-testing.html

 

[Laroxe]

As I said above, there's "risk" in everything, and the word is often thrown around in a handwavey way -- the title of the video is pretty provocative and while they identify the risks, they make no attempt to quantify them or analyze them in context vs other similar risks.

I'd be curious to know if a false-positive DNA match has ever happened. I found one link that suggests a certain test that is pretty prone to false-positives (1 in 700something) once got a guy wrongfully convicted in Taiwan, but that's all I can find. A more common test, though, has a theoretical false positive rate of about 1 in a billion. That's lower than the odds of winning the Powerball and I would hope some basic police work would prevent most such false positives from going anywhere. E.G., I'd hope they wouldn't arrest me for a murder in 1981 in California given that I was a 5 year old living in Pennsylvania at the time.

Is this really an actual worry people have?

This is real, but I wonder what the actual "worry" is, and I'm not sure if it is calculable. If my brother (I don't have a brother that I'm aware of) kills someone and my dad's DNA pops up as a familial match, police will investigate me and my brother. This could include knocking on my door and questioning me. This is new, but is not any different from classic police work. If someone steals my car and uses it to commit a crime, the police will question me about it. If I'm at a restaurant and someone steals money from the cash register but they aren't sure who, I'd expect the police to track me down from my credit card slip and question me. Police (detective) work is all about making connections. The dna family search is just another entry-point. If our culture has people so afraid of the police that they are afraid of being one of the dead-ends in an investigative thread, that's a problem that doesn't have anything to do with DNA.

This "risk" is entirely hypothetical, but paradoxically probably the bigger "worry". The odds of this happening tomorrow(without recourse) are exactly zero since it is currently not legal to do this. The real "worry" I think should be a law being passed that makes it legal. It would be tough to calculate the odds of this ever happening, but I suppose if people want to worry about it, they're entitled to.

I'm not much of a worrier.

Just some notes about part of your post, which may be interesting.
Most of the problems in using DNA evidence are not really about the test itself but the way in which samples are obtained, interpreted and used. In fact attempting to quantify the likely accuracy is now discouraged, remember O.J.'s lawyers ran rings around the DNA experts based on their estimates, even suggesting they were using them to mislead the court - and his arguments were generally sound.

One particularly funny example of problems is

http://content.time.com/time/world/article/0,8599,1888126,00.html

There are lots of issues around the use of DNA evidence that cause problems, these are discussed in a

A European collaboration produced a report that can be downloaded (free)

http://senseaboutscience.org/activities/making-sense-of-forensic-genetics/
There was an extensive review, published as a book called “The Evaluation of Forensic DNA Evidence.” which I believe is all available online. This links to one of the chapters

https://www.ncbi.nlm.nih.gov/books/NBK232607/