How Accurate are Mail-in DNA Tests? 

This channel is underrated but desperately needed.

Mine said “Dude, the only way you could be whiter is if you were Icelandic.”

I read a fantastic article once about a woman who took one of these tests and got results she was certain were wrong. She took it again and got the same wrong results. At that point she started to dig into it farther. After a lot of twists and turns, she wound up discovering that her father had accidentally been switched at birth, totally unbeknownst to anyone! So he (and she by extension) had always grown up thinking he had a certain heritage, but in fact he was related to a completely different family.

Knowing how inaccurate they can be on ancestry, I'd love to know more about how useful DNA tests are for finding out about health risks/predispositions to certain behaviors etc. Avoiding genetic pitfalls is the main reason I've considered doing something like 23 & me.

You should start doing medical deep dives. There's been a few and they're super interesting.

I appreciate this. I think this topic could use a lot more attention. Those tests represent a great opportunity for people to augment their understanding of their own past, but the way they have been marketed and represented in the media has done just the opposite. It's confused people, and messed with their understanding of their past. As a fan of this show, I would love it, if you could do a video explaining to people in detail, just how few of their ancestors are represented in their DNA. I've encountered a lot of people who think these tests are definitive. When I explain that you only carry the DNA of a tiny fraction of your ancestors people still seem to think that a majority of their background would be present in those tests. Armed with a more realistic understanding of how these tests work, I think people could use this information more effectively.

Great Video!

I'd be interested in the results of some kind of actual study here. There are two layers of possible errors. One is the generation of raw results (SNPs in this example, or full sequence data for services that offer this). This video didn't really get into this. How many differences were there in the RAW data between these identical twins in this example? That would tell you a lot about whether the error is a result of random errors in the test measurements (a small number of differences here and there), or a sample mix-up (completely different results). The other layer is in the interpretation. This could be amplified by the random errors. If one of those random errors happens to be in the one base mutation that can cause sickle cell disease, then boom, one person is diagnosed with this condition and the other is not. However, a lot of the interpretation issues weren't really even touched in this video and that is just our general understanding of the relationship between genetic markers and diseases. Some of these are well-known, and many are probabilistic/associative in nature. This is probably the largest area where you need to be careful in interpreting results. This video also doesn't distinguish between SNP testing and exome or full sequence testing. The latter is starting to become affordable - I just ordered a full sequence test from Veritas for what it originally cost me to have an SNP test done from 23andme. Exome testing has been more affordable for a little while. I'm a biochemist by training so I have a reasonable understanding about the limitations of these tests. Many consumers would not. However, not all errors are created equally either, so I'm not sure I like treatments which just talk about "errors" without more nuance.

Oh my God, same. I was so excited to get my ancestry DNA results, only to be told I was almost entirely Eastern European Jewish, which highlighted most of Europe, part of Asia, and the Middle East. Thanks, that really narrowed it down!

A great channel

My daughter bought me a 23 and me DNA test kit I also found the results very general. I can relate to what your saying 100 percent.

But you didn't really answer the question. How accurate are they? We should get a follow-up on how to interpret the data we get back.

In our case - we did 23andMe and there were surprises.But having worked in a genomic lab I'm familiar enough with the process. What I do note, as more and more people have their DNA run the results help to refine the data that 23andMe has. It's fascinating.

I think he's trying to tell us he's of Eastern European Jewish decent.

I've been considering getting a dna test to screen for medical issues( I don't care about ancestry) as I'm adopted and can't rely on family history to highlight potential health problems. Does anyone have any experience with medical mail order DNA tests.

I tried 23andme, but they failed to extract any DNA twice and just gave me a refund. Apparently this is a common issue.

My mom recently took one of these, and it told her pretty much the same things she knew. But it also linked her to her cousin who had taken it... and it said her cousin was partly English or Irish. Which is weird because as far as anyone knows, she has the same background as my mom (and you, I guess): Eastern European Jewish. Funny thing is that her husband is very Irish, so we keep making jokes that maybe she kissed him too deeply before taking the test swab XD

There is no link below??? LIES

Can they find radioactive spider DNA?

I'd be interested in a broader sample and the results statistically. Using anecdotal evidence to support or refute the test results wouldn't be in line with the theme of this channel. How many tests of twins come back with large discrepancies? If the results of %90 of the twin tests are very similar, and only %10 is way off, could we not say that there is a margin of error that must be considered?

So are the mistakes just in terms of placing your ancestry or do they also mistake people for relatives when they aren't?

The way these commercial tests work is they look at the relationship between geography and presence of certain genetic mutations. The computer algorithms are based on how prevalent certain mutations are in an area and the size of the area(s) to which a mutation is common. I had a professor in college who was into this stuff. She showed us some really cool maps showing how certain mutations traveled around the world with well-known migration patterns. I passed them on to my uncle who likes classical genealogy, and his current theory is that as a family of dark-haired Irish people, we’re descended from Sephardic Jews who fled from Spain to Ireland in the 1500s. Too bad these companies engage in such in untrustworthy practices, because I’d really like to see how a DNA test lines up with his work

But do you drive a Volvo and like ABBA? Could be that small percentage of Swedish DNA.

I look forward to understand how microbiome (bacteria) affects DNA expression

I had my son's DNA tested via ancestry.com. I had no idea whatsoever whether any relatives had previously done so. The results matched a few of my cousins, as well as a few of my wife's cousins. The DNA test was accurate. Period. The critique in this video pertains to the ethnicity database. Since that's based in statistical probabilities, confidence can never attain 100%. There will always be statistical outliers. And given the sample size....

Of course there's an ad for a DNA testing company before this video

Hi can uou help me to find my generations

I didn’t see the link to the Vox article but for anyone else who want’s to read it: www.vox.com/platform/amp/science-and-health/2019/1/28/18194560/ancestry-dna-23-me-myheritage-science-explainer

Do a video on arsenic in rice.

Lmao I had the exact same results as you. And it's exactly what I was expecting.

Where’s the link to the article

What about mail in paternity tests? How common are false negatives when parents are a different race? If that even makes a difference.

Lithuanian! Me too!

People should send in their DNA twice or more times under different names. See if the same company gives the same person different results.

Off topic comment! How do humans acquire gut bacteria? Is the DNA involved in the specifics bacteria that each of us has? How different are the gut bacteria between my family or anyone else, or if our ancestors and us? What is needed for DNA to begin building proteins? How

you are eastern european? i thought carroll is irish.

Here's the link that was supposed to be in the description. www.vox.com/science-and-health/2019/1/28/18194560/ancestry-dna-23-me-myheritage-science-explainer

How about a video on why anyone would be stupid enough to volunteer medical PII to an insecure service which isnt covered by HIPPA and outright states that they are going to sell and otherwise misuse your information?

I think the spirit of this episode is great, but I broadly disagree with many of the numbers illustrated. I think 0.1% error rate is extremely high and only true for poor amplifications. Experience in my own work has show high fidelity amplification reactions have an error rate far below the 0.1% threshold . Additionally, given how these tests are run, using chip binding, the signal from a single error would be broadly drowned out by the non-error DNA. Recent studies on ancestry by amplification have shown that ancestry can be largely identified with less than 5000 sites, which is more than an order of magnitude less than those used by most commercial companies. I think skepticism of ancestry by DNA is good and I don't think it is the authority on ancestry. I also think it works infinitely better for those of Caucasian ancestry, but I think the broad flaws in DNA ancestry reporting are due more to historical recording and assumptions scientists make about genetic origins, than the tests themselves.

So basically I should bot waste money?

They are very accurate

💙

More concerned about the fact that the companies are allowed to patent whatever they find in your DNA than I am about the test's accuracy. To my mind, the test is just the decoy to get you to give them something they can use. In any case, hey Aaron, when did your family immigrate to the US? I'm just wondering which political oppression your family was running away from. There was a lot of it, and most of it was from the Tzar, but even if things got better it's not like things were awesome past 1917 either, and *especially* not from '33 to '45, so I'm just curious.

Considering mine said I was mostly Ashkenazi Jewish, it's far from accurate.

Some of them are indeed ACCURATE and it depends on the DNA test you take. Being born in a country means nothing. 23andMe and Ancestry.com do not tell you which clan, tribe or nation you come from. They are approximates at best and the reason for their changes and in inaccuracies. Let's take African countries for example. African countries are all newly invented countries, which did not exist until the mid 1800's in most cases. The second point is that TRIBES in Africa are found all over Africa and some outside of Africa. There's Igbos, Hausa, Fulani, Mende, Nupe, Kanuri, and many, many more other tribes. Let's take the Hausa tribe for example. Hausa tribe are located in Niger, Northern Nigeria, Cameroon, Ghana, Chad, Eritrea, Congo, Gabon, Algeria and even Saudi Arabia. There's over 120 million Hausa people in the world all over Africa and the Middle East. Saying you are from one of those countries means nothing in terms of telling you who you are. BUT, identifying the tribe tells you exactly who you are. You either have DNA from the Hausa tribe or you do not. African Ancestry.com tells you TRIBES specifically. You need to know your CLAN, TRIBE, or NATION, not country, because people move, tribes and nations do not change.

LOCK IT IN 😂

I know those DNA tests were a joke.

If you're 99% Eastern European Jew, why are you talking like a 100% Jersey shore Italian good fellah?

So...DNA is bull...oh. 😂

Meow

Das war ein juden?

Pff... Everyone knows that Homo Sapiens are from Africa! So I do not need to do a genetic test to know where you from. (Assuming you are a Homo Sapiens of course. I only assume that due to how many whom frequent this RU-vid channel are.)

Or maybe they're just making it up. So inaccurate it's hard to tell.

So is there a tried and true DNA test then with no mistakes or as little as possible? From a hospital perhaps? idk but I'd like to get mine done but it seems pointless doing it through these services if it could be bogus information, waste of money.

This channel is underrated but desperately needed.