Half my chromosomal DNA comes from my mother, half from my father. You might be thinking of grandparents. For each chromosome pair (other than XY) I get one from my father and one from my mother with certainty. However, of the one I get from my mother there is a 50% chance she got it from my maternal grandfather and 50% she got it from my maternal grandmother. The chromosome I got from my father could similarly be from my paternal grandfather or grandmother.
I’m a mixed bag and have always known it, but wanted details my parents either could not or would not provide. I considered both Ancestry’s DNA test and 23 and Me, and went with 23 and Me.
I am reported to be:
44.6% broadly Northeast African/Arabian Peninsula (including: Eritrea, Sudan, Libya, Ethiopia, Egypt, Saudi Arabia, and Yemen)
10.5% broadly Northwest African (including: Morocco, Algeria, and Tunisia)
20.8% broadly Northern European (including: Norway, Denmark, and Netherlands)
11.4% French / German
8.7% British / Irish
2.4% Iberian peninsula
1.1% East Asian
.5% Unassigned
It all makes sense to me. I am satisfied with my results.
I’ve done several of these dna tests & 23 & Me is the only one that asked what I thought my ancestry was. However it was optional, so you didn’t have to fill that portion out.
Note in that case a relative had uploaded the raw data to GEDmatch, which is an independent site and they opted in to voluntarily sharing data with the public.
That site changed their rules, and there is a question as to how it would work with 23andme but in this case there were no protections at all for that data.
And wouldn’t that be a Good Thing?
GEDmatch did not change their rules, at least not to prohibit law enforcement use. They did make the potential law enforcement use clearer and may have specified limits on it:
Message on top of main page:
Terms of service detail:
1000 Genomes is a “public catalogue of human variation and genotype data” and is often used as the basis for ancestry and other work in scientific studies. The subjects are all well characterized to be from the ethnicity/population they are assigned. They don’t just find some random person in Spain, and call that the Iberian ideal.
Looking at the map on the link above, you can see there are lots of regions not covered by 1000 Genomes, and any of these genotyping companies will have their own reference samples they use to fill in those gaps.
And yes, 50% of your DNA comes from each parent (assuming no chromosomal abnormalities). But the 50% you share with sibs, 25% with grandparents, 12.5% with first cousins, etc. are just averages. You may just by chance be related more or less than 50% to your sibs. Also, the 25% of grandma’s DNA that you have may be a different 25% than what your brother has, which explains why he is such an idiot.
I have few problems with these companies, which haven’t mentioned, yet (at least, I don’t think so).
First, many of these companies try to sell you on telling you your ethnicity. Sorry, your DNA isn’t going to tell you your ethnicity. It can tell you things about your ancestry, but your ethnicity consists of the foods you like, the music you like, what holidays you observe. That is, the things about your heritage that are specifically NOT related to your ancestry, but to your social situations. OK, I understand that “ethnic” has become euphemistic for “racial” and in that sense, it might apply. I abhor euphemisms, though, specifically because they have a way of making changes to the meanings of words that do not follow any logical path.
Second, while your specific DNA doesn’t have much value, the statistical value of a database of human DNA profiles is very large and becoming more valuable the more we learn about genetics. That is, while the company that collects this information probably will not use it for profit at your detriment, they will most assuredly use it for profit at SOMEBODY’s detriment. By getting you to pay for the testing, but retaining the results for their profit seems a bit dishonest, or at least unethical, to me. Particularly when the results you get (what you pay for) are basically worthless (they admit as much), but what they get has the potential to be incredibly valuable.
Somewhere on these forums, I think in the family secrets thread. A DNA test had some unexpected results and Grandpa admitted that someone had jumped the fence to be with Grandma before they got married. He married her anyway.
I see these DNA tests like those genealogy websites, dig far enough and darken a few blurred lines, and “Congratulations, you’re descended from royalty!”. Now, would you like to a buy a book and a plaque that shows that lineage?
I can see comparing the ethnicity results to that, but mtDNA, YDNA and “family finder” comparisons are actually connecting people with their biological parents, solving genealogy brick walls and firming up well documented pedigrees. Of course sometimes they also reveal family secrets and upend someones pet genealogy theory, but not having found any errors in the 20th century in my own tree I’m willing to live with that. 
My entire family and neighborhood could be traced back to southern central Italy, in two small towns, where people made an exodus together, came through two US immigration centers (Phila and NY) and settled in the same 1 mile square radius of Philadelphia.
No one cracked 88% Italian.
Most are in low 80-percent range.
Some are in the 60-percent.
North African came back on all tests from 1 to 21%
Asian 1-3% came back on most.
A nearby neighborhood of Sicilians and self-proclaimed mobsters (wannabes), after tiring of seeing African blood of over 15-20%, had such turmoil that families have been destroyed and you’ll be outcast if you get a DNA test or speak of a DNA test again… because people… oh… SIGH.
Sicilians needed a DNA test to tell them they have African/North African DNA. #seriously
.
While the African component shouldn’t be surprising with known invasions; this is a good example of how racial theories relate to DNA tests and why caution is needed. DNA does not test heritage and while it is unlikely in the above if you have a mutation that simply isn’t in the dataset that randomly happened to be added to one population, but is far more wide spread than the reference population happened to capture by pure luck the assignment will be very very wrong.
The division of people into races based on political boundaries is taxonomically invalid at the level of the individual. It may provide hints at the population wide migrations but is not as specific as these companies sell to you (but clarify in the fine print).
Some mutations that were assumed to be purely Eurasian as an example were found in Africa just last year, and we don’t have a lot of DNA information from Africa so expect to see more of these cases soon.
http://science.sciencemag.org/content/early/2017/10/11/science.aan8433
“A little learning is a dangerous thing.”
The last one was probably far more common than we realize - or will realize until anyone really tries to crunch DNA numbers - if its even possible. When babies were born on farms and farms were miles apart from each other - and pregnant women didn’t go out in public, and adoption carries stigma as well as unwed motherhood, there is incentive - and it isn’t difficult - to pass off your cousin’s child out of wedlock as your own.
It is family lore at least one of my great aunts is a several times removed cousin. “Go visit the cousins on the farm” and leave your baby there…
Mismatched mothers might be hard to find from genetic data, though, because in most of those cases, the actual and purported mothers are probably going to be close relatives. By contrast, if a woman has an affair, it might be but probably isn’t with a close relative of her husband.
If the purported mother is really a grandmother or aunt, then it is going to be very obvious very quickly. A mother and child should share 50% of their DNA, but if the number is closer to 25%, then that will be noticed.
However, if two “siblings” are really mother and child, then they’ll share 50% of their DNA, and we won’t ever notice something is odd. That is actually much more likely to happen, because in most genetic studies DNA is not collected from the parents. It is statistically more powerful to add another subject than to add a subject’s parent, and given finite money, it will be spent to increase the power of the study. Collecting DNA (and other information) from multiple siblings is very common.
There is more genetic variation within Africa than outside of it. I (a person of European descent) am closer related to a person of Chinese descent than two “black” people might be to each other. (That probably isn’t true if the black people are African Americans descended from slaves, but may be true if the two black people are native Africans from different regions.) For lots of both innocent and historically racist reasons most genetic work has been done on people of European descent. Fortunately there has been a big push over the last 10 or so years to change this. As the genetics of more African groups are characterized, we’re bound to learn really interesting things about the history of Homo sapiens, and things that can be directly applied to all of humanity today.
Oh, I quite get it.
The "DNA industry’’ – if that’s a thing – comes from a position of perceived authority, and with graphs, data, fancy packaging and soft-focus commercials, it’s bound to be a mess at times for families and individuals. A co-worker currently has a problem in his family as he and his sisters have results that are so vastly different, they are now arguing with their mother and contacting their aunts for the truth about their father, their heritage, etc.
Political boundaries: Some boundaries can drive genetics, such as the Alps, and bodies of water that restrict movement can create some local patterns among people who have ancestors that stayed put. It would take more time than I have to look at what seems to be a limitless number of variables.
I tell my family that my ancestors were a bunch of illiterate, poor, isolated farmers, who came here because they were being starved out by their own people in the northern, industrialized parts of Italy, and I simply drive home the fact that DNA doesn’t prove you are part of anything special, because it was Italians starving Italians that mattered most. Details of that aside, it makes them temper their enthusiasm for blind allegiance to people within the lines on a map.
As someone who worked for a company in Milan, I do have anecdotes to tell them about how people in Northern Italy look down at the peons as far south as Naples and especially Sicily. My family have come to learn more about the little towns we came from, and just baking cookies this month gives me a chance to remind them that the cookies and food we make this month are about our little Southern Italian towns and clans, and really not about the romanticized version of Italy.
100% Italian!!! was a bragging right. Now, they are just proud of how their ancestors carved out a new life here.
.
.
For anyone interested, here (warning: pdf)is a pretty detailed description of how Ancestry.com determines ethnicity (their term) from an individual’s DNA. Bottom line: people of single-origin ancestry won’t get a 100% match. Even Italians on their reference panels averaged less than 90% Italian. This does not mean that their tests are bullshit, it just means that there is going to be a substantial amount of statistical error. For those 100% Italians who come back at 90%, that remaining 10% will show as something else.
I’m pretty sure my results would come back as 100% American mutt.