Ancestry Tests
in Purgatory
Has anyone done an ancestry test? I'm considering one but I have a few questions.
1) How accurate is it?
2) Does discovering one's ancestry make someone rethink their overall identity?
3) Anything else I should be aware of?
1) How accurate is it?
2) Does discovering one's ancestry make someone rethink their overall identity?
3) Anything else I should be aware of?
Comments
Personally I don't think it would make a huge difference to know my genetic heritage, unless I suppose something came back really unexpected. (E.g., neither of my parents are Italian, so if my genetic heritage turned out to be exactly half-Italian it might raise some awkward questions.) I'm basically a mongrel of various northern European ethnicities, as far as I know. I would be surprised to find out that I had significant genetic heritage from, well, a long list of places, but I don't think it would have any effect on my sense of who I am.
This is really rather dodgy. Here's the reasons.
These places use reference pools to judge your DNA against. That is, they went out (or paid others who had already gone out) to get DNA from places judged to be as pure as possible, genetically speaking--which is to say, not very, in a lot of cases) populations that had been on the same piece of land for generations. How many generations is a sticking point, because a) you're relying on their self-report, and b) if you go back far enough in history, EVERYBODY comes from migrants. Who hail from somewhere else... ultimately, most likely Africa, for everybody.
So you have to take it largely on faith that this particular reference population, which your DNA happens to match, stands for something meaningful about your historical ancestors. And that might not be so. All the more because they aren't testing ALL of your DNA against ALL of theirs, but only selected bits. And you yourself don't possess the DNA of all your ancestors, but only a very selected bit, mostly shared with your most recent ancestors--so the Cherokee great-great-great-grandfather's DNA might be entirely written over by the time it gets to your generation. Or not, as chance has it.
There's also the fact that if you have unusual DNA (read: non-European), you will probably not get much detail, because their reference pools skew heavily heavily European. Good luck if you're, say, Burmese. You may get back a notation like East Asian or even "Asian, general" which is a fat lot of good.
You will also find that your results change if you check them every few years, as they continue to add more of these reference pools. Which is a bit disconcerting. So consider it basically a parlor game.
I suspect the kits are more useful for tracing relatives of the past few generations.
My cousin discovered he was not really a part of the family. He took one and saw no connection with us. He called me and the first thing I assured him was as far as we were concerned he was a part of the family. I then encouraged him to reach out to some of the names that came back as first cousins. Turns out they want nothing to do with him. He found out from his mother (with whom he is estranged--she abandoned him as a kid) that she had been hitchhiking in Montana and a trucker picked her up and one thing lead to another. (It was the era of free love.)
In my research, I had to redo two branches. Ancestry DNA had indicated certain people were my direct descendants, but checking with a couple of other genealogy sites the dates did not add up. I had to retrace a couple of names. In both cases, Ancestry suggested I was a descendant of one brother, but it was another brother. Once I identified the correct brother, things pretty well fell into place.
There is one roadblock I am still dealing with. It is my great-grandmother on my maternal grandfather's side. I have a death certificate that lists her name, but the line disappears. My Mother says she was an orphan. I might not get around it.
I'm also querying why it's so interesting. Having several adopted family members who're as much my family as those who weren't adopted and perhaps because who I'm fonder of. Also, as we've 4 different races represented both by marriage and adoption, it has been commented among us that any preoccupation with genetic ancestry contains probable worrying aspects of racism. But this may reflect having (blood!) relatives who were Nazis, and the rejection of anything related by the 2 generations since.
I've a family tree that purports to go to 1000AD which was cooked up to prove pure Aryan blood. So it's all a crock to me.
Firstly, it was because I knew so little about my mother (who was adopted) that I wanted to see if there were any familial links out there. As it turned out, the DNA test didn't help at all but I subsequently discovered (almost by accident) that my mother had a half brother she knew nothing about and so that I have first cousins, who can tell me more about my mother's birth-mother.
Secondly, my father was pretty vague about his side of the family, so I wanted to see if there were any links that might help me. Nothing so far.
Thirdly, I am researching a few people who might be distant relatives and I had a very faint hope that a DNA test might throw up some clues.
Finally, there is always the possibility that the results might have thrown up something truly weird (like a strong Eastern Europe or Middle East connection). I know a few people who this has happened to.
Of course, there's always a chance that it will produce results you really don't want (like discovering that your father isn't your father). If you don't want to face something like that, don't do it.
And I'm a bit cautious considering that in all likelihood, their research database on East Asian populations might be limited, so I question the reliability.
Still I am curious if it's possible that this test might be useful, hence my curiosity.
You can probably google for the companies with tests most likely to be helpful to people with Asian ancestry. Almost certainly some will be trying to branch out into this area.
If you would otherwise be interested, seek out a company that will allow you to add more coverage later without a new round of exams and disclosures. Then pick up a tiny policy--say, enough for funeral expenses.
There was one thing in my results that would have been very surprising if I didn't already know about it. The first result I got was about my unexpected half-brother, who had looked us up and made contact. He's my mother's eldest son, and none of us had any clue that he existed. But he was born in Kansas, which is an open records state. All he had to do was send off for his birth certificate.
I hate family secrets. My mother had to guard that one her whole life. I think I can see evidence of the difficulty it caused her when I recall certain comments she made over the years.
I don't know how much to trust the ethnicity guesses you get from these companies. Company #1 gave me a rundown that repeated what I already knew - basically Britain and Ireland. Company #2 added a head-scratcher. According to them I've got about 2% Inuit in my background.
American Indian wouldn't be surprising - I've got lots of ancestors in colonial New England, and somebody might well have married somebody who had that connection. But Inuit? I can't figure that one out at all. I figure I'll just wait and see if they have second thoughts about it.
All that said, I wouldn't take it for the reasons outlined by @Marsupial and @Sober Preacher's Kid. I'll take my genetic test, if ever, when it becomes medically relevant, regardless of my curiosity.
It's enabled me to confirm one great great grandfather, which I could not have done through conventional research.
I will never stop being amazed by the connections it throws up. That said, it would be less interesting if I hadn't already had a paper family history to map the results onto.
They do update their findings from time to time as more people enter the database.
So even when you're dead and gone to heaven, you're still bothered by the kids for help?
[Genuine question] How can you disprove something like this? I'd have thought all it takes is one person in the chain who (putting in bluntly) shags everything in sight and suddenly you're in a whole other branch of the tree. Over the course of 400 years, say a dozen generations or so, the chances of the official father not being the actual father, or an undocumented adoption, get significant. Is there some way of correcting for this?
I’ve considered ancestry tests but I expect it would be very unexciting results.
There are some trees on Ancestry that are so obviously wrong, that the person putting them up clearly hasn't looked at them; just copied and pasted - in any tree if someone allegedly died aged 130, or had their first child aged three, or if they appear to have been skipping back and forward between Canada in Scotland in the early C19th, commonsense should suggest a problem.
But Ancestry DNA tests are not affected by crap trees. DNA is what it is.
I've been researching my family history for decades. My PhD was a study of social mobility which involved following hundreds of families through successive census returns, and creating small family trees. I have a lot of experience of primary sources.
I have always known that the paper records don't necessarily reflect reality. Much of the excitement of DNA for me was that the DNA has confirmed my research. I met someone years ago through family history - she and I are descended from a couple who married in 1795. When I saw her name as a distant cousin DNA match, I cried.
I'd probably join in - I did most of mine the old fashioned way, so I'm waaay behind on what can be done online.
I'm with ST on Fawkes Cat's post. Brilliant!
I've taken such a test and from my experience it's pretty accurate. My results matched up with family lore and correctly identified several known relations who had taken the same company's test, including their degree of relatedness. The caveats are that I have a genetic background that's somewhat over-represented in DNA databases.
Not really, but I had a vague but comprehensive notion of my ancestry before taking the test, which didn't contain any major surprises. There was one minor thing which had long been speculated on my mother's side which the test confirmed, but nothing earthshaking.
The caveat here is that you should be prepared in case there is something surprising. I remember hearing an interview with someone who, in the early days of commercial genetic ancestry tests, decided to use her family as a "control group" to verify the accuracy of the tests. The results came out pretty much as expected except for one of her uncles. It turned out the test was accurate, just his paternity wasn't what everyone thought.
Other complications include things like white supremacists discovering they've got African (or other non-white) ancestry.
As others have noted, sometimes law enforcement will use DNA test results to track down criminals. You probably shouldn't take the test if you've committed any crimes for which DNA evidence might exist. A conversation along these lines:
On a technical note (I am not a geneticist, so take this for what it's worth), this isn't as much of a problem as it's being portrayed. A lot of the human genome is standardized and doesn't vary that much regionally. Hox genes, for example, dictate the form and sequence of embryonic development, meaning mutations in these genes are often fatal. They're so stable the same Hox genes are found in all bilateral animals (humans, birds, insects, etc.) and a few radials as well. Genes that control critical things aren't as subject to mutation because such mutations are often fatal and thus strongly selected against. If everyone has the same Hox genes (or a very limited number of variations) those genes are useless for determining human ancestry. They'd work to show you weren't some kind of plant, but presumably you already knew that.
Of course they don't bother with the bits that all humans have. But they can't be bothering with a lot of the bits only SOME humans have, either, because that would be time-and cost-prohibitive. Let's say (pulling an number out of thin air) that there are 15,000 genes that differentiate me from ... your sister, or anybody's sister (so we don't get into it about sex). They can't be looking at all 15,000 of those, it would take ages and cost the earth. From my reading, they have certain target "bits" and use those. And I have not yet come across any explanation how how they choose the target bits--whether it's for maximal variation, for convenience, or some other reason or combo of reasons.
So we're looking at (very roughly) 6 million base pairs to compare, which is a helluva lot of comparison to ask of a place that's charging you 99$ for a Christmas present. I really doubt they're looking at all or even ten percent.
Here's an explainer from Vox.
And remember, some of these SNPs don't make a measurable difference in how a gene is expressed. For example, AAA and AAG both code for Lysine.
There’s also the fact that while you get 50% of your DNA from each parent it’s not necessarily an evenly distributed 50%. You may have 3 billion base pairs (per the aforementioned Vox article), but they’re grouped together on 46 chromosomes and assigned to you en masse. Think of it like reaching into two bags of tokens with 46 tokens in each and drawing out 23 from each. Even if the two bags have 23 ‘Irish’ tokens between the two of them there’s no guarantee that you’ll draw exactly 11 or 12 of them. You might end up with 20 of them. Or 3.
But the bigger problem is the on-going blither they send. They discover a match. Both my first wife and Kuruman come from Families Of Distinction.™ So I am constantly getting references to their famous friggin' ancestors. And really I don't need to know that my ex-wife's third cousin's uncle's third wife was Murgatrotd Frigging Flobbdegoo.
An American sent my Dad our tree back to 1700-ish (we have an odd surname) though we were able to correct a couple of points about the modern bits - so maybe not all is correct. It reads like a human geography lesson - generations born in shit on a farm in one small village in the sticks in Essex, leave for East London over two generations around 1850, mostly live in the same road until the 1960s, dissipate all over the place since. I'd love some Murgatrotd Frigging Flobbdegoo, but that's centuries of Essex vanity showing. If I had some I'd name them on a brass-effect plate screwed to my uPVC porch columns
But it does mean it is probably not worth me having DNA ancestry as I am almost certainly also going to be solidly Saxon, with possibly a little Viking.
Sounds like my family, only substitute South for East London. And does go some way to explain the (to us) shocking snobbery of one long-since ‘late’ relative over those who work in ‘trade.’ I bet you they’d have loved to have had a Murgatrod Double-Barrel in the tree.