# Please help me interpret my 23andMe ancestry results **URL:** \ **Category:** In My Humble Opinion\ **Created:** [December 9, 2018, 2:55pm UTC](https://boards.straightdope.com/t/please-help-me-interpret-my-23andme-ancestry-results/825849 "2018-12-09T14:55:50Z")\ **Posts on this page:** 20\ **Page:** 4
**Author:** ![Pardel-Lux](https://sea3.discourse-cdn.com/straightdope/user_avatar/boards.straightdope.com/pardel-lux/32/20182_2.png) [@Pardel-Lux](https://boards.straightdope.com/u/Pardel-Lux)\ **Post date:** [December 13, 2018, 8:39pm UTC](https://boards.straightdope.com/t/please-help-me-interpret-my-23andme-ancestry-results/825849/61 "2018-12-13T20:39:10Z")
> [@Lamoral](#): > > " OK, so “they” have my DNA, what the hell is anyone going to do with it? Why should I be concerned? If you ever commit a serious crime, murder or so, you will have to be very cautious with the traces you leave. And they will get your relatives easily if they commit a crime. Which may be good or bad, depending on your relatives. I know what I think of most of mine 😃 ---
**Author:** ![Nawth\_Chucka](https://avatars.discourse-cdn.com/v4/letter/n/c57346/32.png) [@Nawth\_Chucka](https://boards.straightdope.com/u/Nawth_Chucka)\ **Post date:** [December 13, 2018, 10:55pm UTC](https://boards.straightdope.com/t/please-help-me-interpret-my-23andme-ancestry-results/825849/62 "2018-12-13T22:55:39Z")
The 23andMe results are much more subjective than they’re presented, and recombination doesn’t make that any less confusing. For example, my mom’s always told us she has a Native American ancestor; my results show no NA whatsoever while my (same parents) brother’s show 4% NA. If we relied on my results alone we’d say that eliminates a Cherokee ancestor but w/ my brother’s results we can’t. The sole exact number my brother and I share is our Neanderthal comparison; greater than 74% of the tested population. I wonder which explains my lactose intolerance? 😃 ---
**Author:** ![Isamu](https://avatars.discourse-cdn.com/v4/letter/i/7c8e57/32.png) [@Isamu](https://boards.straightdope.com/u/Isamu)\ **Post date:** [December 14, 2018, 6:45am UTC](https://boards.straightdope.com/t/please-help-me-interpret-my-23andme-ancestry-results/825849/63 "2018-12-14T06:45:50Z")
You’re part centaur. ---
**Author:** ![Sam\_Lowry](https://avatars.discourse-cdn.com/v4/letter/s/e95f7d/32.png) [@Sam\_Lowry](https://boards.straightdope.com/u/Sam_Lowry)\ **Post date:** [December 14, 2018, 3:44pm UTC](https://boards.straightdope.com/t/please-help-me-interpret-my-23andme-ancestry-results/825849/64 "2018-12-14T15:44:33Z")
Someone please correct me if I’m wildly wrong, but from what I’ve understood reading about this is that your DNA is like a novel, but the DNA testing companies can’t read the absolute whole thing, and have to make educated guesses on what they can’t read. If someone was reading Jane Eyre and there were chunks missing, they might be able to accurately summarize the story or they might get major things wrong depending on what chunks of the book they are missing. And this also explains why different DNA companies could give you different results, or why close relatives could get results that are more different than expected, because the company is looking at different parts of the book. ---
**Author:** ![lingyi](https://avatars.discourse-cdn.com/v4/letter/l/d2c977/32.png) [@lingyi](https://boards.straightdope.com/u/lingyi)\ **Post date:** [December 14, 2018, 4:42pm UTC](https://boards.straightdope.com/t/please-help-me-interpret-my-23andme-ancestry-results/825849/65 "2018-12-14T16:42:45Z")
Even identical twins who were thought to have the same exact DNA have been proven they don’t. [https://www.biotechniques.com/omics/not-so-identical-twins/](https://www.biotechniques.com/omics/not-so-identical-twins/) ---
**Author:** ![Sailboat](https://sea3.discourse-cdn.com/straightdope/user_avatar/boards.straightdope.com/sailboat/32/461_2.png) [@Sailboat](https://boards.straightdope.com/u/Sailboat)\ **Post date:** [December 14, 2018, 5:48pm UTC](https://boards.straightdope.com/t/please-help-me-interpret-my-23andme-ancestry-results/825849/66 "2018-12-14T17:48:28Z")
Great thread title / username combo! Looking at your percentages, I see a microcosm of World War I (Western Front, at least) ---
**Author:** ![echoreply](https://sea3.discourse-cdn.com/straightdope/user_avatar/boards.straightdope.com/echoreply/32/3641_2.png) [@echoreply](https://boards.straightdope.com/u/echoreply)\ **Post date:** [December 14, 2018, 6:29pm UTC](https://boards.straightdope.com/t/please-help-me-interpret-my-23andme-ancestry-results/825849/67 "2018-12-14T18:29:44Z")
> [@Sam\_Lowry](#): > > Someone please correct me if I’m wildly wrong, but from what I’ve understood reading about this is that your DNA is like a novel, but the DNA testing companies can’t read the absolute whole thing, and have to make educated guesses on what they can’t read. If someone was reading Jane Eyre and there were chunks missing, they might be able to accurately summarize the story or they might get major things wrong depending on what chunks of the book they are missing. > > And this also explains why different DNA companies could give you different results, or why close relatives could get results that are more different than expected, because the company is looking at different parts of the book. That’s a bit of a strained analogy. In the case of ancestry, I’d say it’s less trying to summarize the story, than trying to identify the book. Your DNA is a whole jumble of shredded pages. The different companies have different methods and libraries they are matching it to. One comes back and says we think your jumble is a mashup of _Jane Eyre_ with _War and Peace_. The other company comes back and reports _War and Peace_, but they don’t have _Jane Eyre_ in their library, so they say it sort of kinda looks like _Wuthering Heights_ to them, and your all, that’s not even the right sister. > [@lingyi](#): > > Even identical twins who were thought to have the same exact DNA have been proven they don’t. [https://www.biotechniques.com/omics/not-so-identical-twins/](https://www.biotechniques.com/omics/not-so-identical-twins/) Take it one step further, different tissue in your body don’t have the same DNA. For example, sequencing cancer cells and non-cancerous cells can reveal what mutations might be causing the cancer. That’s all super interesting, but is way zoomed in too far for discussing ancestry. To use another strained analogy, it is like telling somebody how to get to your house, and saying that Google Maps shows your driveway 6 inches north of where it actually is located. It’s not going to make a bit of difference for the direction finding use case, but if the surveyor gets your driveway off by 6 inches, that might be a big deal. ---
**Author:** ![XT](https://sea3.discourse-cdn.com/straightdope/user_avatar/boards.straightdope.com/xt/32/456_2.png) [@XT](https://boards.straightdope.com/u/XT)\ **Post date:** [December 14, 2018, 7:58pm UTC](https://boards.straightdope.com/t/please-help-me-interpret-my-23andme-ancestry-results/825849/68 "2018-12-14T19:58:11Z")
> [@What\_the](#): > > My brothers and I are trying to convince my 90 year old dad to do it. He’s the last one on his side left and our mom has passed too. > > One of his problems with it, is that my results are not as Italian as he thinks they should be. Below are the pertinent results. I’m wondering if there is something about the testing itself or Italy as a country to come from that would explain it and convince him that it would support his proud Italian heritage. > > Italian … 1.5% > French/German… 38.9 > British/Irish …30.3 > Broadly NW European…17.8 > Broadly Southern Euro…5.3 > Broadly European …2.5 > Iberian…2.2 We had the same issue wrt family history and lore. My dad, for instance, expected our results to show Greek ancestry since one of our family stories is about a Greek ancestor who was trying to come to the US and ended up in Mexico. When we did the test, however, there was zero Greek. What there was that was unexpected, was some Southern Italian. My WAG was that the ‘Greek’ ancestor (assuming such a person existed) was either of Italian stock that colonized Greece sometime in the past or was actually from Italy (which would be pretty much the same thing for most Mexicans in the 1800’s). Most of the rest of my results made sense. Lots of Native American, some German/French (both of who were heavily in Mexico at various times), and of course Iberian from the Spanish (though happily, when my 2.0 results came out it turns out they were Portuguese) and then some Eastern European/Russian from my mother’s side. The most exotic thing I had was some Turkish/ME blood and, off the wall and out of now where, 2% from India (gods know where THAT came from). Mainly the results were what we thought, but my dad was seriously bummed about the Greek thing. I think I’ve been able to explain that away with the whole Roman Empire colonized Greece in the past part, but it was definitely a discontinuity. Not surprising since, unlike seemingly a lot of Americans (or Europeans), we don’t actually know much more than oral family history for my dad’s side of the family. ---
**Author:** ![rat\_avatar](https://sea3.discourse-cdn.com/straightdope/user_avatar/boards.straightdope.com/rat_avatar/32/255_2.png) [@rat\_avatar](https://boards.straightdope.com/u/rat_avatar)\ **Post date:** [December 14, 2018, 8:52pm UTC](https://boards.straightdope.com/t/please-help-me-interpret-my-23andme-ancestry-results/825849/69 "2018-12-14T20:52:07Z")
> [@XT](#): > > We had the same issue wrt family history and lore. My dad, for instance, expected our results to show Greek ancestry since one of our family stories is about a Greek ancestor who was trying to come to the US and ended up in Mexico. When we did the test, however, there was zero Greek. What there was that was unexpected, was some Southern Italian. My WAG was that the ‘Greek’ ancestor (assuming such a person existed) was either of Italian stock that colonized Greece sometime in the past or was actually from Italy (which would be pretty much the same thing for most Mexicans in the 1800’s). > > Most of the rest of my results made sense. Lots of Native American, some German/French (both of who were heavily in Mexico at various times), and of course Iberian from the Spanish (though happily, when my 2.0 results came out it turns out they were Portuguese) and then some Eastern European/Russian from my mother’s side. The most exotic thing I had was some Turkish/ME blood and, off the wall and out of now where, 2% from India (gods know where THAT came from). Mainly the results were what we thought, but my dad was seriously bummed about the Greek thing. I think I’ve been able to explain that away with the whole Roman Empire colonized Greece in the past part, but it was definitely a discontinuity. Not surprising since, unlike seemingly a lot of Americans (or Europeans), we don’t actually know much more than oral family history for my dad’s side of the family. Scientifically there is no such thing as an “Italian” or “Greek” gene. Italy and Greece are close enough that it is actually _more likely_ that the DNA service you used simply didn’t have a reference population from Greece with those markers. While the general public’s response to ancestry testing has been enthusiastic, scientists response is barely luke warm. Most of the data behind these ancestry assignments are proprietary, in this study 92% of the alleles were found in more than one region and half were found in all 7 regions they tested. > **[popstructSupp.pdf](https://rosenberglab.stanford.edu/supplements/popstructSupp.pdf)** > > 205.15 KB From a scientific biological perspective separate racial or ethnic groups _do not_ actually exist. Even when they find region-specific alleles they typically only occur in a tiny fraction of that population. For heritage genetics is simply a field of _probabilities_ and not of _hard facts_. It can tell you where other people who happen to share an allele were from. With that information you an _infer_ that others may be from the same area, but is unable to say your father has “zero Greek”. In sociology an _ethnic group_ is a group of people who share a common heritage, culture, and/or language. In the U.S., _ethnicity_ is often used to refer to _race_. Heritage, culture, language and even “race” are not based on biology or passed via DNA. DNA can be a useful tool for _inferring_ a connection with these groups but is is not hard evidence for or against connections. I want to be clear that I am not suggesting that DNA can’t be a useful tool, but in cases like yours it is extremely important to remember the limitations of this method. While it is a bummer, for heritage, DNA can provide _hints_ but it will not provide _answers_ without additional reliable evidence. ---
**Author:** ![XT](https://sea3.discourse-cdn.com/straightdope/user_avatar/boards.straightdope.com/xt/32/456_2.png) [@XT](https://boards.straightdope.com/u/XT)\ **Post date:** [December 14, 2018, 9:08pm UTC](https://boards.straightdope.com/t/please-help-me-interpret-my-23andme-ancestry-results/825849/70 "2018-12-14T21:08:19Z")
> [@rat\_avatar](#): > > Scientifically there is no such thing as an “Italian” or “Greek” gene. Italy and Greece are close enough that it is actually _more likely_ that the DNA service you used simply didn’t have a reference population from Greece with those markers. > > While the general public’s response to ancestry testing has been enthusiastic, scientists response is barely luke warm. Most of the data behind these ancestry assignments are proprietary, in this study 92% of the alleles were found in more than one region and half were found in all 7 regions they tested. > > [https://rosenberglab.stanford.edu/supplements/popstructSupp.pdf](https://rosenberglab.stanford.edu/supplements/popstructSupp.pdf) > > From a scientific biological perspective separate racial or ethnic groups _do not_ actually exist. Even when they find region-specific alleles they typically only occur in a tiny fraction of that population. For heritage genetics is simply a field of _probabilities_ and not of _hard facts_. It can tell you where other people who happen to share an allele were from. With that information you an _infer_ that others may be from the same area, but is unable to say your father has “zero Greek”. > > In sociology an _ethnic group_ is a group of people who share a common heritage, culture, and/or language. In the U.S., _ethnicity_ is often used to refer to _race_. > > Heritage, culture, language and even “race” are not based on biology or passed via DNA. DNA can be a useful tool for _inferring_ a connection with these groups but is is not hard evidence for or against connections. > > I want to be clear that I am not suggesting that DNA can’t be a useful tool, but in cases like yours it is extremely important to remember the limitations of this method. > > While it is a bummer, for heritage, DNA can provide _hints_ but it will not provide _answers_ without additional reliable evidence. And Ancestry (which is the service we used instead of 23 and me) discussed a lot of what you are saying, once you dig in past the dashboard. And, intellectually, I get it. But to my family, they are looking at the pretty colors and percentages on a world map, and the colors on said map had no Greek, but instead had Southern Italian. Since a lot of the other results are in line with their expectations (i.e. we knew we had a lot of Native American and that it was from the South West…and that’s exactly what we saw, South West US and Northern Mexico and partially into Central Mexico), it tends to reinforce that there is something to this to someone who doesn’t have a clear idea of even what a gene is. It also makes explaining things like why the results might not conform to family history…challenging. Personally, I found the results interesting, but a lot of my family was much more into it since, as I noted, so little of our own actual family history outside of oral stories is known. Literally going back beyond my own grandmother on my fathers side is like pre-history, as it’s only some stories with zero evidence, so something like DNA at least gives something to folks like that. I have to say that, on my mothers side, the results are pretty much what is documented, so while it might not be scientific, it conforms pretty well to the records we have…i.e. part of my mothers family comes from Eastern Europe, and that’s basically what Ancestry said as well. ---
**Author:** ![echoreply](https://sea3.discourse-cdn.com/straightdope/user_avatar/boards.straightdope.com/echoreply/32/3641_2.png) [@echoreply](https://boards.straightdope.com/u/echoreply)\ **Post date:** [December 14, 2018, 10:22pm UTC](https://boards.straightdope.com/t/please-help-me-interpret-my-23andme-ancestry-results/825849/71 "2018-12-14T22:22:44Z")
> [@rat\_avatar](#): > > From a scientific biological perspective separate racial or ethnic groups _do not_ actually exist. Even when they find region-specific alleles they typically only occur in a tiny fraction of that population. For heritage genetics is simply a field of _probabilities_ and not of _hard facts_. It can tell you where other people who happen to share an allele were from. With that information you an _infer_ that others may be from the same area, but is unable to say your father has “zero Greek”. Saying that race and ethnic groups don’t exist is a bit of an oversimplification. What I would say is the more complex statement that in humans there is genetic variation in which there are clusters of similarity. People can be grouped by these similarities. At a very narrow level, these similarities will identify first and second degree relatives. At a much broader level these similarities can identify groups with similar ancestry. Among scientists, these ancestry groups are often referred to as “races,” but the mapping of these genetically determined racial groups to traditional racial groups is poor at best. For example, an African American may actually have 70% European genetic ancestry, but out in the world will be called “black” just the same as a person with 100% African genetic ancestry. And when we do genetic studies, we usually use principal components, which is a statistical test that converts the similarity in DNA between subjects into values. The values that individuals get on a particular component will cluster in relation to genetic ancestry. This doesn’t give clean answers, and experience and rules of thumb go into much of the determination of where to draw lines separating clusters. Depending on the analysis being done, it might be appropriate to do the analysis separately for different clusters, or it might only be necessary to take into account subjects’ principal components scores as part of the analysis. The problem is, that there are many pitfalls in genetic analyses if the clustering of similarity—racial groups—is ignored. \*I’m saying all of that with my genetics researcher hat on, and mean it all in as connotation free way as possible from \*\*_a white privileged hetero cis male_ seat. I’m not in any way trying to minimize the damage that has been caused by racism and prejudice. I can deny that culturally assigned race is a good reason to hate people, and at the same time accept that there are observable and important differences between groups of people. \* ---
**Author:** ![rat\_avatar](https://sea3.discourse-cdn.com/straightdope/user_avatar/boards.straightdope.com/rat_avatar/32/255_2.png) [@rat\_avatar](https://boards.straightdope.com/u/rat_avatar)\ **Post date:** [December 14, 2018, 11:13pm UTC](https://boards.straightdope.com/t/please-help-me-interpret-my-23andme-ancestry-results/825849/72 "2018-12-14T23:13:45Z")
> [@echoreply](#): > > Saying that race and ethnic groups don’t exist is a bit of an oversimplification. What I would say is the more complex statement that in humans there is genetic variation in which there are clusters of similarity. People can be grouped by these similarities. At a very narrow level, these similarities will identify first and second degree relatives. At a much broader level these similarities can identify groups with similar ancestry. Among scientists, these ancestry groups are often referred to as “races,” but the mapping of these genetically determined racial groups to traditional racial groups is poor at best. > > For example, an African American may actually have 70% European genetic ancestry, but out in the world will be called “black” just the same as a person with 100% African genetic ancestry. > > And when we do genetic studies, we usually use principal components, which is a statistical test that converts the similarity in DNA between subjects into values. The values that individuals get on a particular component will cluster in relation to genetic ancestry. This doesn’t give clean answers, and experience and rules of thumb go into much of the determination of where to draw lines separating clusters. Depending on the analysis being done, it might be appropriate to do the analysis separately for different clusters, or it might only be necessary to take into account subjects’ principal components scores as part of the analysis. The problem is, that there are many pitfalls in genetic analyses if the clustering of similarity—racial groups—is ignored. > > \*I’m saying all of that with my genetics researcher hat on, and mean it all in as connotation free way as possible from \*\*_a white privileged hetero cis male_ seat. I’m not in any way trying to minimize the damage that has been caused by racism and prejudice. I can deny that culturally assigned race is a good reason to hate people, and at the same time accept that there are observable and important differences between groups of people. > \* Feel free to point me to a cite showing a scientific biological basis for race and ethnic groups and that is based on DNA. Note I am not talking about finding markers and estimating the _probability_ that someone may match a data-set, but that there are actual genetic delineations and not just fuzzy associations with convenient correlations. Plus I never said that race and ethnic groups don’t exist, they are _social constructs_ but so is money. The effects of these social constructs are very very real. ---
**Author:** ![rat\_avatar](https://sea3.discourse-cdn.com/straightdope/user_avatar/boards.straightdope.com/rat_avatar/32/255_2.png) [@rat\_avatar](https://boards.straightdope.com/u/rat_avatar)\ **Post date:** [December 14, 2018, 11:30pm UTC](https://boards.straightdope.com/t/please-help-me-interpret-my-23andme-ancestry-results/825849/73 "2018-12-14T23:30:39Z")
I know this is IMHO, but to help people out who may be interested in this: > **[Patterns of Human Diversity, within and among Continents, Inferred from...](https://genome.cshlp.org/content/12/4/602.long)** > > An international, peer-reviewed genome sciences journal featuring outstanding original research that offers novel insights into the biology of all organisms > [@](#): > > These results suggest that, at random biallelic loci, there is little evidence, if any, of a clear subdivision of humans into biologically defined groups. Here is a cite on where this does matter, but it will never be “Greek or Italian” which are useless labels > [@](#): > > In other words, for epidemiological purposes, _similar genetics_ matter and _ancestry labels_ do not matter. > **[Different differences: The use of ‘genetic ancestry’ versus race in...](https://www.ncbi.nlm.nih.gov/pmc/articles/PMC3124377/)** > > This article presents findings from our ethnographic research on biomedical scientists’ studies of human genetic variation and common complex disease. We examine the socio-material work involved in genome-wide association studies (GWAS) and discuss... Note this quote from that paper that directly related to the issues posters have communicated in this thread. > [@](#): > > Although genetic ancestry is not race, these ancestry categories are often read as race and ethnicity, especially when the ancestry labels match national, tribal, and other designations that have socio-cultural meaning and carry political implications. ---
**Author:** ![echoreply](https://sea3.discourse-cdn.com/straightdope/user_avatar/boards.straightdope.com/echoreply/32/3641_2.png) [@echoreply](https://boards.straightdope.com/u/echoreply)\ **Post date:** [December 15, 2018, 4:06am UTC](https://boards.straightdope.com/t/please-help-me-interpret-my-23andme-ancestry-results/825849/74 "2018-12-15T04:06:13Z")
> [@rat\_avatar](#): > > Feel free to point me to a cite showing a scientific biological basis for race and ethnic groups and that is based on DNA. Note I am not talking about finding markers and estimating the _probability_ that someone may match a data-set, but that there are actual genetic delineations and not just fuzzy associations with convenient correlations. > > Plus I never said that race and ethnic groups don’t exist, they are _social constructs_ but so is money. The effects of these social constructs are very very real. I’m not quite sure what you are asking. What is the difference between a genetic delineation and a fuzzy association? These things are all probabilities. Sometimes it is an easy call, lots of people sort into very distinct clusters. Other times it is much harder, what about the edges where clusters overlap? If you want strict rules and determinism, then stay away from genetics, and probably all of biology. There are differences in DNA between people. These differences are not completely random. People with more similar ancestry will probably have more similar DNA, and people with more similar DNA will probably have more similar ancestry. For better or worse, we in the field call those groups races and ethnic groups. Maybe that doesn’t meet your definition of a biological race, but that is the term used. You asks for cites, but that is a bit nonsensical. Show me a genetic study that _doesn’t_ mention race and ethnicity and I’ll show you something that should never have been published (and yes, they exist). Any groups who are reproductively isolated by mountains, oceans, distance, culture, or whatever will become genetically distinct over time. They are still all humans and can interbreed with other groups given the opportunity (and will). So these reproductively isolated groups will all be living in a similar region, where maybe they share a language, culture, and other things. So all of the sudden we have a social construct—this group of people who share a language and culture. Through random genetic drift, a founder effect, or whatever, they also share some DNA. Now we have a pattern of DNA that maps on to a social construct. There are also lots of people who are a mixture of different groups. The term we use is admixture. Most people who identify as African American are a mix of African and European genetic backgrounds. Most people who identify as Hispanic/Latino are varying mixtures of European, Native American, and African genetic backgrounds. All of this shows the wonderful complexity and diversity of humanity. ---
**Author:** ![naita](https://sea3.discourse-cdn.com/straightdope/user_avatar/boards.straightdope.com/naita/32/5862_2.png) [@naita](https://boards.straightdope.com/u/naita)\ **Post date:** [December 15, 2018, 3:21pm UTC](https://boards.straightdope.com/t/please-help-me-interpret-my-23andme-ancestry-results/825849/75 "2018-12-15T15:21:25Z")
> [@Nawth\_Chucka](#): > > The 23andMe results are much more subjective than they’re presented, and recombination doesn’t make that any less confusing. For example, my mom’s always told us she has a Native American ancestor; my results show no NA whatsoever while my (same parents) brother’s show 4% NA. If we relied on my results alone we’d say that eliminates a Cherokee ancestor but w/ my brother’s results we can’t. > The sole exact number my brother and I share is our Neanderthal comparison; greater than 74% of the tested population. I wonder which explains my lactose intolerance? 😃 “If we relied on my results alone we’d say that eliminates a Cherokee ancestor”. If you said that you’d be wrong, even if the test was absolutely accurate down to single digit percentages. At that distance it is entirely possible to have 0% DNA from a specific ancestor, so a test can confirm, but eliminate, ancestry. ---
**Author:** ![naita](https://sea3.discourse-cdn.com/straightdope/user_avatar/boards.straightdope.com/naita/32/5862_2.png) [@naita](https://boards.straightdope.com/u/naita)\ **Post date:** [December 15, 2018, 3:42pm UTC](https://boards.straightdope.com/t/please-help-me-interpret-my-23andme-ancestry-results/825849/76 "2018-12-15T15:42:48Z")
> [@Sam\_Lowry](#): > > Someone please correct me if I’m wildly wrong, but from what I’ve understood reading about this is that your DNA is like a novel, but the DNA testing companies can’t read the absolute whole thing, and have to make educated guesses on what they can’t read. If someone was reading Jane Eyre and there were chunks missing, they might be able to accurately summarize the story or they might get major things wrong depending on what chunks of the book they are missing. > > And this also explains why different DNA companies could give you different results, or why close relatives could get results that are more different than expected, because the company is looking at different parts of the book. If we stick to the novel analogy everyone has a copy of Jane Eyre, a really long one, 3 billion letters. There are 10 million known typos with varying frequencies and DNA-tests look at 500k to 1 million of these, but not necessarily the same, although the overlap between all, except the most recent 23andMe\*, is way above 50%. The different results between companies are more about it being unscientific to pick people today with a few generations of ancestry in a region and assume they are particularly representative of the population going back centuries. If we could test everyone in, say 1500, there would probably be more clumping than we see today, but there’d be clumps within national borders and there would be clumps spanning borders. The population today isn’t the result of perfect mixing of these clumps within national borders or “ethnicities”, and I was just picking 1500 as a random point anyway. To take one concrete example. Ancestry “updated” their estimates to split up Scandinavian to Norwegian and Swedish. It is reasonable to assume that there’s a continuum of variation in Norway and Sweden, with the far West and South of Norway being the most different from the far East and South of Sweden, the southern border region having more of a mix, and the far north having a mix as well, but different, with a heavy element of Sami and Finn. So when Ancestry slaps the label Norwegian or Swedish on these bits of DNA, what does that actually mean? It certainly doesn’t mean “that Norwegian” DNA can’t have existed entirely within the Swedish borders for many generations. ---
**Author:** ![XT](https://sea3.discourse-cdn.com/straightdope/user_avatar/boards.straightdope.com/xt/32/456_2.png) [@XT](https://boards.straightdope.com/u/XT)\ **Post date:** [December 15, 2018, 8:43pm UTC](https://boards.straightdope.com/t/please-help-me-interpret-my-23andme-ancestry-results/825849/77 "2018-12-15T20:43:10Z")
> [@rat\_avatar](#): > > I know this is IMHO, but to help people out who may be interested in this: > > [https://genome.cshlp.org/content/12/4/602.long](https://genome.cshlp.org/content/12/4/602.long) > > Here is a cite on where this does matter, but it will never be “Greek or Italian” which are useless labels > > [Different differences: The use of ‘genetic ancestry’ versus race in biomedical human genetic research - PMC](https://www.ncbi.nlm.nih.gov/pmc/articles/PMC3124377/) > > Note this quote from that paper that directly related to the issues posters have communicated in this thread. You seem to be fixated on the labels for some odd reason, as if you think most folks on this board don’t get the whole race thing. Since you seem fixated on it, I’ll clarify (again) that the Ancestry results gave broad regions of people, and my particular results as with most of my immediate family gave southern Italy as the region of origin for some of our ancestors (about 9%), but didn’t label us ‘Greek’ or ‘Italian’. There was, of course, several regions in, you know, Greece, that we apparently didn’t have any percentage of ancestry from. Now, if you want to make the case that the regional indicators are complete bullshit, feel free to do so. I’m all ears as I don’t pretend to be even well versed on this sort of thing. But if you want to go on and on about the labels I put on things for convenience because folks will understand those labels as broad indicators of regional populations then you are basically just picking nits for your own pleasure. ---
**Author:** ![septimus](https://sea3.discourse-cdn.com/straightdope/user_avatar/boards.straightdope.com/septimus/32/410_2.png) [@septimus](https://boards.straightdope.com/u/septimus)\ **Post date:** [December 16, 2018, 7:56am UTC](https://boards.straightdope.com/t/please-help-me-interpret-my-23andme-ancestry-results/825849/78 "2018-12-16T07:56:44Z")
Does human genetic variation form clusters or clines? **echoreply** points out, I think, that the distinction between clusters and clines is fuzzy; and that Principal component analysis is the way forward. For visualization it’s common to plot the 2 or 3 highest variance components on [a diagram like this one.](http://www.molecularecologist.com/wp-content/uploads/2014/05/Tishkoff2009_fig02.png) African Americans and Native Americans are definitely in different ‘clusters,’ but the distinction between East Asians and Native Americans might be more like a ‘cline.’ It might be interesting to compare that diagram with one for a species with an accepted division into distinct subspecies. ---
**Author:** ![septimus](https://sea3.discourse-cdn.com/straightdope/user_avatar/boards.straightdope.com/septimus/32/410_2.png) [@septimus](https://boards.straightdope.com/u/septimus)\ **Post date:** [December 16, 2018, 8:49am UTC](https://boards.straightdope.com/t/please-help-me-interpret-my-23andme-ancestry-results/825849/79 "2018-12-16T08:49:00Z")
> [@septimus](#): > > It might be interesting to compare that diagram with one for a species with an accepted division into distinct subspecies. [Here’s a paper discussing and diagramming genetic diversity among species and subspecies of great apes.](https://www.nature.com/articles/nature12228) It might be an interesting paper; I’ve not waded through it. ---
**Author:** ![rat\_avatar](https://sea3.discourse-cdn.com/straightdope/user_avatar/boards.straightdope.com/rat_avatar/32/255_2.png) [@rat\_avatar](https://boards.straightdope.com/u/rat_avatar)\ **Post date:** [December 16, 2018, 9:30pm UTC](https://boards.straightdope.com/t/please-help-me-interpret-my-23andme-ancestry-results/825849/80 "2018-12-16T21:30:24Z")
> [@XT](#): > > You seem to be fixated on the labels for some odd reason, as if you think most folks on this board don’t get the whole race thing. Since you seem fixated on it, I’ll clarify (again) that the Ancestry results gave broad regions of people, and my particular results as with most of my immediate family gave southern Italy as the region of origin for some of our ancestors (about 9%), but didn’t label us ‘Greek’ or ‘Italian’. There was, of course, several regions in, you know, Greece, that we apparently didn’t have any percentage of ancestry from. > > Now, if you want to make the case that the regional indicators are complete bullshit, feel free to do so. I’m all ears as I don’t pretend to be even well versed on this sort of thing. But if you want to go on and on about the labels I put on things for convenience because folks will understand those labels as broad indicators of regional populations then you are basically just picking nits for your own pleasure. Here is a good page related to this, the potential for psychological, social, and political harm is high without realizing the limitations. > **["Plug and Play" Genetics, Racial Migrations and Human History](https://www.scientificamerican.com/blog/observations/plug-and-play-genetics-racial-migrations-and-human-history/)** > > Genetically profiling individuals is now cheap and easy, but writing our species’ global history still calls for a lot of hard work Regional indicators are complete bullshit with any length of time even going back to the 17th or 18th century. The number individuals who could have contributed to your genetics is so large and their possible genetic contribution so small (0.098% for 10 generations back), it is useless for any claims about who they were in person. Regional indicators have some use for a snapshot of a population at one point in time but outside of selling services to consumers their value is minimal from a scientific point of view related to “heritage” which isn’t even a genetic trait. [Previous page](https://boards.straightdope.com/t/please-help-me-interpret-my-23andme-ancestry-results/825849.md?page=3) [Next page](https://boards.straightdope.com/t/please-help-me-interpret-my-23andme-ancestry-results/825849.md?page=5)