irish ethnicity percentage meaning
hi! I have recently received my mother's DNA test results and according to the groupping , she appears as 58% Iberian and 31% Irish (scotish and welsh). I am a bit surprised as we don't have (as far as I know) close recent ancestors of this ethnicity. My question is: by having a 30% , if I understand it right, it means that either a parent or grandparent should have been 100% Irish, is this right? I know we have a distant ancestor (9 generations) who was English, but I wouldn't expect that much DNA in my mother's result. Can somebody help me understand what 30% means genetically speaking? thanks a lot for your help!
Best Answer
-
If I may suggest, you will do better asking your question on a DNA-specific platform. There are many good DNA support groups on various social media platforms.
And depending on where you tested, your results can vary widely.
As Judy G Russell says, It's not soup yet. The results are just estimates.
1
Answers
-
You will want to watch this from this year's RootsTech: https://www.familysearch.org/en/rootstech/session/you-can-do-the-dna-2-get-your-best-ethnicity-estimate
Pay particular attention to the section on Reference Populations.
2 -
Oy. Problem with the lecture that Gordon linked: it states a complete falsehood as its basic tenet. (I only watched as far as 0:54, so I don't actually know where she went with it. I'm an identical twin, hence my immediate rejection of watching any further.)
I agree with Judy Russell; my usual phrasing is that DNA "ethnicity" estimates have "entertainment value only".
The testing companies all look at only a small portion of people's DNA, because most of it is exactly the same in all people. (In fact, a lot of it is the same in all animals, and some of it is the same in all living things. Yes, we share DNA with trees.) The choice of which genes to look at is based on experience, and varies slightly by company.
If the sequence of molecules in one of the variable spots on a chromosome is the same between two people, it can basically have two explanations: either it happened by chance, or they inherited it from a common ancestor. The longer the sequence, or the more such sequences two people share, the less likely the first explanation becomes. When comparing two specific people, such pattern-matching is statistically quite good at identifying likely relatives.
The "ethnicity" estimates, on the other hand, are statistically not so reliable. They're basically saying that if everyone in Group A had Pattern 1, but nobody in Group B had it, then Pattern 1 must be characteristic of Group A, and therefore, anyone who has Pattern 1 must share ancestors with Group A. The problem is, all of the underlying assumptions may be false. Pattern 1 may also occur in Group B, just not in the people they happened to test, or it may occasionally occur in Group A (including all of the people they happened to test), but it could actually originate from and be characteristic of Group C, which the company didn't include in its reference populations. Add in migration, intermarriage, and politics, and you end up with estimates that are suitable only for entertainment.
Or, to put it another way: geography is not genetic.
1 -
"cocktail party conversation pieces" as Judy Russell often says.
Ethnicity is why many people test, but the high matches with other testers are far more important.
0 -
Yeah, I heard that "no one has your exact DNA" and immediately thought of identical twins but I'm sure that was just a slip of the tongue to not slide in that exception. However the rest of the lecture was very interesting, particularly her explanation of the behind the scenes fiddling the companies do to assign those estimates.
Very intriguing was her discussion of the fact that if a person has DNA that does not match any reference group, no company is ever going to say "we don't know what you are." They are going to take the closest available reference group and say you belong to it.
One of the audience questions at the end that shows how fuzzy these estimates are was from a woman who asked why one company said she was 15% Scandinavian, one said 35%, and one said 0%. I'll let everyone listen to the answer themselves.
1 -
thank you all for bringing some light on my ignorance :) This is really helpful. I will dig in the family tree and also contact My heritage (there is where the DNA results are) to see if they can explain more. I guess that, since I am missing some branches of the tree, I thought that there could be a direct relation with this result, but as you have correctly pointed, the most reliable info is the direct matches, not the group.
thanks again for the good advise and good job in the comunity!
0