My sister recently had her DNA tested by ancestry.com, along with her daughters and our dads. Question is , the daughter came up with no Scandinavian DNA…my sister has 34% Scandinavian, 26% Eastern European, and 40% Central European DNA…why would it come up zero percent for her daughter? BTW, all three came up as being close relatives, such as parent or sibling…
The scientist in me is setting off bells that the testing method is not entirely accurate and flawed.
I haven’t studied this so am not an expert in consumer grade DNA testing such as this; but I don’t think the ability to tell ancestory with great accuracy through consumer DNA testing products is all that good.
Pretty much as evidenced by your sisters experience
What fraction of the genome could be identified as Scandinavian-specific?
What ancestry did the daughter’s father show?
First of all, it’s possible for one parent to have genes from a particular heritage that the child lacks. A child gets only one chromosome from each chromosome pair from each parent, so it could be that the mother’s Scandinavian genes were all on chromosomes that didn’t get passed to that particular kid.
Second, scientific measurements of any sort always have error bars on them. A reputable report shouldn’t just say something like “34%”; it should say something like “34% ± 10%” (or maybe ±5% or whatever; I don’t know the details of the tests).
Third, we don’t know how many genes the lab looks at to associate with each heritage. If it’s a small number, then they’ll vary more wildly.
Yeah, a human has IIRC 48 chromosomes, in pairs, half from the mother, half from the father. Assuming the Scandinavian heritage of the mother is from one side ofher family, then there are no duplicates, so there is no pair of chromosomes where both of the pair are Scandinavian (Even if both the mother’s parentss have a bit, it’s possible she ended up with no duplicates). It’s not impossible for the daughter to have none, assuming the father has none.
Each chromosome constitutes about 2%. 34% Scandinavian (Sc) indicates about 17 out of 48 are Sc. The odds of getting NONE of those is incredibly high, assuming they actually test for all 48 chromosomes, not just a few choice markers on a small selection of the chromosomes.
I think that most likely they are looking for markers, not actually decoding the entire genome of those individuals. Suppose there are 6 known markers that, when seen together, indicate a person has Scandinavian ancestry. It’s possible that the child gets none of those markers, or let’s say they get 2. The analyst might need to see at lest 4 or 5, together, to state there is evidence of Scandinavian ancestry.
And just to be clear, your sister doesn’t have “34% Scandinavian DNA”. We’re all 99.9% the same, so 99.9% of our DNA all comes from the same place (Africa). That little bit that differentiates us from each other might have bits that are seen mostly in Scandinavians. It’s not hard to see how we might not get that tiny bit from one of our parents.
Natl. Geographic has a similar project: https://genographic.nationalgeographic.com/
Or you can study the DNA of total strangers: Creepy or Cool? Portraits Derived From the DNA in Hair and Gum Found in Public Places | Science| Smithsonian Magazine
Or even if he does have some. She could have just gotten the non-Scandinavian part of each of his pairs, too.
Maybe md2000 got two more wholesale.
Ancestry.com is pretty vague on their site about what they’re testing and how these things are linked, but it looks like the percentages are confidence intervals, not necessarily a percentage composition. i.e. 34% means 34% chance of Scandinavian heritage, not that you are 34% Scandinavian. (In fact, it means a 66% chance of NOT being Scandinavian.)
From here: AncestryDNA FAQ
Or he doubled Down.
I saw what you did there.
The way the question and supporting information is stated by the o.p. is so misleading that it isn’t really possible to answer the question. I don’t know if this is just poor reporting on the part of the company that assayed the genome or if the o.p. is being very imprecise about how the report stated the percentage, but what genealogical DNAtests do is to look for unique genetic ‘markers’ (haplotypes) in the code which are associated with specific genetic populations. These are largely confined to haplotypes found in mitochondrial (non-nuclear) DNA and the Y chromosome as these are passed directly from the mother-to-all-children and father-to-son directly without recombination, and thus are very stable and consistent from generation to generation, unlike the non-Y chromosome nuclear DNA, the specific chromosomes of which are inherited only roughly half of the time; that is, during meosis, of a matching pair of chromosomes one combined into the genome of the gamate and the other is thrown away. There are also single nucleotide polymorphisms (SNP) which are found in the recombined code with many, many repeats such that even with the random chance of inheriting any particular marker the sheer volume assures their presence in a certain haplogroup.
So, say a haplogroup known to have emerged from the Scandinavian region carries a certain marker; the prevalence of different markers will be compared to known databases of this haplogroup in order to determine the relative likelihood that a test subject emerged from this group and (in more sophisticated analysis) how long or with what degree of divergence, e.g. if there are (hypothetically speaking) twenty known markers for a person who has direct ancestry from a group and the test subject only has ten of these markers shown then they will have experienced significant divergence from that group, either due to loss by mutation and gene shift, or (in the case of SNP) many generations of recombination with non-carriers of the marker.
Note, this doesn’t mean that x percentage of your genes are “Scandinavian” or whatever. As John Mace notes, an overwhelming amount of the genome is common amongst all human beings, and indeed, amongst all mammals. All these numbers mean is a probability that you have derived some ancestry from that particular group and the genetic distance from which you have diverged from it, which is used to estimate the portion of ancestors identified as coming from specific haplogropus. And frankly, a lot of it is statistical guesswork rather than discrete accounting, so variability even with family (and frankly, from test to test) is hardly surprising.
It is worth noting that while we have historically categorized “Blacks” of African descent into one large genetic grouping (‘race’ or ethnicity or whathaveyou) the reality is that the amount of genetic variability in haplogroups geographically identified with African populations is vastly larger than the entire rest of the worlds population combined. This fact is surprising to the average person but should be obvious to anyone familiar with breeding and the so-called ‘Founder effect’; the populations that left Africa and settled Asia, Europe, and later the Americas and Australia were small populations which were often isolated or enjoyed limited interaction with populations of signficantly divergent genetic heritage. The initial pool of genes being smaller means that the overall genetic breadth is more restricted, even as such populatoins evolved phenotypes to adapt to varying environmental conditions (lighter skin, tolerance for lactose, casein, and gluten, et cetera) and passed on apparent random mutations such as freckles and green eyes. In fact, the greatest amount of genetic variability of populations outside of Africa may well have come from hybridizing with other subspecies of Homo sapiens such as H. neanderthalensis, H. heidelbergensis, and potentially other as-yet unknown contemporaneous human species.
Stranger
ccarljc, there is no such thing as Scandinavian DNA, or Eastern European DNA, or the rest of the types you mention. If they are telling you there is, they are either lying to you or they are oversimplifying their explanation to such a degree as to make it both false and incoherent. (So it is hardly your fault if you can’t understand it, because what they are saying just does not make sense).
What there may be are certain markers within a persons DNA, that is, certain rare variants of a gene, that are much more common in people of Scandinavian ancestry than amongst others, or more common amongst people of Eastern European ancestry, or whatever. If you find, say a Scandinavian marker in someone’s DNA that tells you that there is a good chance (a probability on which they can put a number) that that person had a Scandinavian ancestor some number of generations back. It is not a certainty, however, because (at least in most cases) just because a certain marker is more common in Scandinavians (or whoever) it does not follow that it never, ever turns up in other parts of the world, in people who have no Scandinavian ancestors at all. Thus there is usually some, small chance that the person with the marker might have gotten it from a non-Scandinavian, and there is also a chance that a person without any distinctively Scandinavian markers, might nevertheless be of largely Scandinavian descent.
I may not be explaining things entirely clearly: I am very tired, and the matter is, indeed, very complicated (even though I am still giving a highly simplified version of how it really works). Believe me, though, there is no such thing as having 34% Scandinavian DNA, nor, for that matter, 0% Scandinavian DNA. This sort of DNA analysis just gives you statistical odds that you may have, or may not have, ancestors from a particular region.There is no such thing a Scandinavian DNA, or Easren European DNA, or the rest of the types you mention.
ETA: I see stranger has explained things far better than I was able to manage,and in far more detail while I was still composing my very inadequate post. I should not try to post while I am dopey with fatigue.