# What Should I Conclude from These DNA Tests? **URL:** \ **Category:** Factual Questions\ **Created:** [December 24, 2017, 10:02pm UTC](https://boards.straightdope.com/t/what-should-i-conclude-from-these-dna-tests/805242 "2017-12-24T22:02:44Z")\ **Posts on this page:** 18\ **Page:** 2
**Author:** ![pkbites](https://sea3.discourse-cdn.com/straightdope/user_avatar/boards.straightdope.com/pkbites/32/204_2.png) [@pkbites](https://boards.straightdope.com/u/pkbites)\ **Post date:** [December 25, 2017, 5:50am UTC](https://boards.straightdope.com/t/what-should-i-conclude-from-these-dna-tests/805242/21 "2017-12-25T05:50:16Z")
> [@Hari\_Seldon](#): > > I read somewhere (sorry no cite) that careful DNA tests in England showed that about 10% of all people didn’t have the father they thought they had. [It happens.](http://boards.straightdope.com/sdmb/showthread.php?t=821464) ---
**Author:** ![Duckster](https://sea3.discourse-cdn.com/straightdope/user_avatar/boards.straightdope.com/duckster/32/1244_2.png) [@Duckster](https://boards.straightdope.com/u/Duckster)\ **Post date:** [December 25, 2017, 10:10pm UTC](https://boards.straightdope.com/t/what-should-i-conclude-from-these-dna-tests/805242/22 "2017-12-25T22:10:18Z")
> [@Hari\_Seldon](#): > > Just for the record, I would not have my DNA tested if it were free. Or if you paid me. Sometimes ignorance should not be fought. [What you’re giving away with those home DNA tests](https://www.nbcnews.com/health/health-news/what-you-re-giving-away-those-home-dna-tests-n824776). ---
**Author:** ![LordHarry](https://avatars.discourse-cdn.com/v4/letter/l/6bbea6/32.png) [@LordHarry](https://boards.straightdope.com/u/LordHarry)\ **Post date:** [December 26, 2017, 1:00am UTC](https://boards.straightdope.com/t/what-should-i-conclude-from-these-dna-tests/805242/23 "2017-12-26T01:00:56Z")
This is good advice, and I will try the GEDmatch. I cannot do Y chromosome analysis, as I do not have a Y chromosome. I am avoiding disclosing surname here, but it is an Italian surname and the entire crew of greatgrandparents and great(2) grandparents on that side are from San Mango Sul Calore, Italy. My grandmother on that side was adopted, so although her parents were also from this area, they don’t “count” in the genetic question. > [@Yeah](#): > > I suggest you upload your DNA data to GEDmatch. It is free and it will give you another set of possible cousins as well as another admixture of ethnicities. > > Based on my experience, you will get only slightly different results from each analysis. For example, one might have “British Isles” while the other one has “English” and “Irish.” And one might have a couple percent South Asian or Eastern Mediterranean while the other doesn’t. You also might find some previously unknown matches. > > Regarding admixture, I would not pay much attention to anything \<5% unless it shows up in more than one analysis. One analysis, but not the other, shows me as “\<2% American Indian” despite the fact that none of my ancestors were in North America before the Civil War. > > Also, while the big DNA companies seem to be on the up and up, I don’t think they do a good job helping you interpret results. I am pretty sure that if you took 100 “purely English” persons and analyzed their DNA, many would be 5%-20% Scandinavian and many would 5%-10% French or German (much higher in the case of the Royal Family). So when you see 2% Scandinavian that does not necessarily mean there was one purebred Scandinavian on holiday in England a number of generations ago. I could be that several of your English ancestors’ very distant ancestors had 5%-20% Scandinavian ancestry. Remember, you can get 25% from one grandparent who was 100% or from two grandparents who were 50% or from four great grandparents who were … > > I have looked carefully through the “2nd-4th” cousin matches for myself and my wife and have never found any previously unknown relatives. These people do not necessarily share a recent ancestor. They could just as easily share multiple bits of DNA from multiple shared distant ancestors. > > My impression, from what you have shared, is that it is possible that your father is not your ancestor. But, where in Italy is he from? What is his surname?. Not every family living for a long time in Italy is 100% Italian. More important, what does your Y-chromosome test show? > > If I were you, I would get my Y-chromosome test to see whom I matched. This reflects exclusively the male line so there is a good chance it will give you a definitive result regarding your paternity. Second, I would upload my DNA findings to GEDmatch and then check GEDmatch for matches and admixture. ---
**Author:** ![md2000](https://avatars.discourse-cdn.com/v4/letter/m/73ab20/32.png) [@md2000](https://boards.straightdope.com/u/md2000)\ **Post date:** [December 26, 2017, 2:53am UTC](https://boards.straightdope.com/t/what-should-i-conclude-from-these-dna-tests/805242/24 "2017-12-26T02:53:30Z")
The term I heard was “pedigree error”. There’s the apocryphal story of the teacher in a lower class London area doing a class on blood types, asking each student to check their blood types with their parents to demonstrate genetics- only to find that 25% of children suffered from pedigree error. (1) BS, how many lower class British from the 1970s or earlier knew their blood type? (2) Even there number seems way too high (3) probably just slander, put-down of the lower class - demonstrate their moral turpitude, and (4) since the more common blood types are very common and recessive usually, the odds that even if 25% were bastards, that 25% would fail a blood match is unlikely - it would probably suggest an error at the level of 40% or more. And (5) I’ve searched for the original source and never found any reference. It’s urban legend material. IIRC I read something about that 10% study which included a lot of adoptees and remarriages, basically 10% of children were not the offspring of both adults in the current household - which is quite plausible in almost any town in the western world given today’s divorce statistics. It didn’t mean hidden parentage. But reading between the lines, it seems like mama had a fling (most likely explanation) either for purpose of expanding the family (if the famial resemblance is not too bad) or for other reasons (if the OP is distinctly different.) The first thing to do would be to find where mama worked or who the close family friends were around that time, if you can. …of half-English heritage. There’s even an ad on TV for I think Ancestry where the person says “I though our family was Italian, turns out we’re eastern European…” Family migrated to Italy a generation or two earlier. IIRC too, when they tested slave and free descendants of Thomas Jefferson (or alleged) some were, some weren’t the same family tree… Suggesting some pedigree errors in the intervening years. Plus, it could have been his father or uncle… ---
**Author:** ![jasg](https://avatars.discourse-cdn.com/v4/letter/j/f0a364/32.png) [@jasg](https://boards.straightdope.com/u/jasg)\ **Post date:** [December 26, 2017, 3:32am UTC](https://boards.straightdope.com/t/what-should-i-conclude-from-these-dna-tests/805242/25 "2017-12-26T03:32:40Z")
The OP might find this article interesting. > **[She thought she was Irish — until a DNA test opened a 100-year-old mystery](https://www.washingtonpost.com/graphics/2017/lifestyle/she-thought-she-was-irish-until-a-dna-test-opened-a-100-year-old-mystery/)** > > How Alice Collins Plebuch’s foray into “recreational genomics” upended a family tree. ---
**Author:** ![Surreal](https://avatars.discourse-cdn.com/v4/letter/s/eada6e/32.png) [@Surreal](https://boards.straightdope.com/u/Surreal)\ **Post date:** [December 26, 2017, 4:44am UTC](https://boards.straightdope.com/t/what-should-i-conclude-from-these-dna-tests/805242/26 "2017-12-26T04:44:07Z")
> [@md2000](#): > > And (5) I’ve searched for the original source and never found any reference. It’s urban legend material. [http://www.childsupportanalysis.co.uk/analysis\_and\_opinion/choices\_and\_behaviours/misattributed\_paternity.htm#mclaren1977](http://www.childsupportanalysis.co.uk/analysis_and_opinion/choices_and_behaviours/misattributed_paternity.htm#mclaren1977) [http://www.childsupportanalysis.co.uk/analysis\_and\_opinion/choices\_and\_behaviours/misattributed\_paternity.htm#philipp1973](http://www.childsupportanalysis.co.uk/analysis_and_opinion/choices_and_behaviours/misattributed_paternity.htm#philipp1973) ---
**Author:** ![Surreal](https://avatars.discourse-cdn.com/v4/letter/s/eada6e/32.png) [@Surreal](https://boards.straightdope.com/u/Surreal)\ **Post date:** [December 26, 2017, 5:17am UTC](https://boards.straightdope.com/t/what-should-i-conclude-from-these-dna-tests/805242/27 "2017-12-26T05:17:09Z")
> [@Hari\_Seldon](#): > > I read somewhere (sorry no cite) that careful DNA tests in England showed that about 10% of all people didn’t have the father they thought they had. [http://www.journals.uchicago.edu/doi/abs/10.1086/504167](http://www.journals.uchicago.edu/doi/abs/10.1086/504167) > [@](#): > > Evolutionary theory predicts that males will provide less parental investment for putative offspring who are unlikely to be their actual offspring. Crossculturally, paternity confidence (a mans assessment of the likelihood that he is the father of a putative child) is positively associated with mens involvement with children and with investment or inheritance from paternal kin. **A survey of 67 studies reporting nonpaternity suggests that for men with high paternity confidence rates of nonpaternity are(excluding studies of unknown methodology) typically 1.9%, substantially less than the typical rates of 10% or higher cited by many researchers.** Further crosscultural investigation of the relationship between paternity and paternity confidence is warranted. ---
**Author:** ![Cazzle](https://sea3.discourse-cdn.com/straightdope/user_avatar/boards.straightdope.com/cazzle/32/75_2.png) [@Cazzle](https://boards.straightdope.com/u/Cazzle)\ **Post date:** [December 26, 2017, 5:20am UTC](https://boards.straightdope.com/t/what-should-i-conclude-from-these-dna-tests/805242/28 "2017-12-26T05:20:15Z")
I wouldn’t bother retesting with My Heritage - I’ve seen lots of posts calling their accuracy into doubt. Specifically, I’ve seen other genealogists calling them out for predicting relationships between tested individuals to be falsely close, where the other sites have all correctly predicted relationships to be more distant. Their ethnicity predictions have also been mentioned to be harder to align with documented ancestry than the other sites. The straight up, simple route is to have a documented relative from your paternal family test, and to contact people with trees containing your paternal ancestors to establish if there’s a documented connection, and if there is then find out if they have tested. Your close family seem to be out of the question, but a more distant relative is less likely to be caught up in whatever unpleasantness has been going on. If you can find a second or third cousin, that would do the trick. Fourth cousins will not match you 50% of the time so not matching a single fourth cousin would be inconclusive. Keep in mind that your documented family tree only goes back so far; the ethinicty estimates go back further. Almost no one living in Ireland, for example, with documented Irish ancestry back hundreds of years would come up as 100% Irish. ---
**Author:** ![Johanna](https://sea3.discourse-cdn.com/straightdope/user_avatar/boards.straightdope.com/johanna/32/8318_2.png) [@Johanna](https://boards.straightdope.com/u/Johanna)\ **Post date:** [December 26, 2017, 2:11pm UTC](https://boards.straightdope.com/t/what-should-i-conclude-from-these-dna-tests/805242/29 "2017-12-26T14:11:03Z")
> [@LordHarry](#): > > This is good advice, and I will try the GEDmatch. I cannot do Y chromosome analysis, as I do not have a Y chromosome. You’re a trans man? You mentioned family transphobia earlier; I can totally relate. > [@](#): > > I am avoiding disclosing surname here, but it is an Italian surname and the entire crew of greatgrandparents and great(2) grandparents on that side are from San Mango Sul Calore, Italy. Deep in the [heart](https://en.wikipedia.org/wiki/San_Mango_sul_Calore) of Campania—you’re of Samnite origins. I’m a Messinese trans woman and I love studying Italian history and prehistory. I just got done reading Elena Ferrante’s Neapolitan novels (_My Brilliant Friend_ etc.), about people from your region. My family gave me grief too. ---
**Author:** ![Smeghead](https://avatars.discourse-cdn.com/v4/letter/s/f1d935/32.png) [@Smeghead](https://boards.straightdope.com/u/Smeghead)\ **Post date:** [December 27, 2017, 5:39pm UTC](https://boards.straightdope.com/t/what-should-i-conclude-from-these-dna-tests/805242/30 "2017-12-27T17:39:17Z")
> [@LordHarry](#): > > So DNA would expect to see 25% Italian, 25% WILD CARD (father’s side) and Mostly English, with some Irish on my mother’s side. Not necessarily. Although exactly\* half of your DNA comes from your father, it is not true that that half must be precisely divided between his parents. It is, in theory, possible for you to have anything between 0% and 50% from any particular grandparent. In practice, the odds are very very good that the true number will be close to 25%, but it’s random selection. As a professional geneticist, I would not take anything from these tests beyond perhaps motivation to get proper forensic work done if you’re really interested. It’s not that there’s anything wrong with the tests themselves - they’re all done at LabCorp, and they’re competent - it’s just that the links between certain markers and certain populations are never rock solid. They’re all based on probability and don’t always reflect exactly what we think they do. I would be really really cautious drawing any conclusions or taking any action based on these reports alone. If nothing else, there’s always the chance that a traveling salesman came through town a hundred years ago and left some strange DNA behind him. ---
**Author:** ![jacobsta811](https://sea3.discourse-cdn.com/straightdope/user_avatar/boards.straightdope.com/jacobsta811/32/2893_2.png) [@jacobsta811](https://boards.straightdope.com/u/jacobsta811)\ **Post date:** [December 27, 2017, 6:45pm UTC](https://boards.straightdope.com/t/what-should-i-conclude-from-these-dna-tests/805242/31 "2017-12-27T18:45:07Z")
I would take the Country origins portion with an enormous, heaping grain of salt. The portion you should look at to determine paternity/non-paternity is the DNA relatives section only. You can and should test at all 3 of the major sites, which are [Ancestry.com](http://Ancestry.com) (which you did), [23andme.com](http://23andme.com), and Family Tree DNA. MyHeritage uses the same chip as one of the three major ones and their DNA matching seems suspect so I wouldn’t use them at all. Get as many relatives as you can to test, and as close a relative as possible. Particularly important if possible are parents, aunt and uncles, and grandparents as those help you filter out relatives by which “side” of the family they are on. ---
**Author:** ![md2000](https://avatars.discourse-cdn.com/v4/letter/m/73ab20/32.png) [@md2000](https://boards.straightdope.com/u/md2000)\ **Post date:** [December 28, 2017, 7:20am UTC](https://boards.straightdope.com/t/what-should-i-conclude-from-these-dna-tests/805242/32 "2017-12-28T07:20:24Z")
> [@Surreal](#): > > [http://www.childsupportanalysis.co.uk/analysis\_and\_opinion/choices\_and\_behaviours/misattributed\_paternity.htm#mclaren1977](http://www.childsupportanalysis.co.uk/analysis_and_opinion/choices_and_behaviours/misattributed_paternity.htm#mclaren1977) > > [http://www.childsupportanalysis.co.uk/analysis\_and\_opinion/choices\_and\_behaviours/misattributed\_paternity.htm#philipp1973](http://www.childsupportanalysis.co.uk/analysis_and_opinion/choices_and_behaviours/misattributed_paternity.htm#philipp1973) yes, notice those are “unpublished” especially the 20% to 30% is sort of peripherally mentioned in another discussion. Note the preliminary remarks - > [@](#): > > “Rates of non-paternity have taken on the character of urban folk tales … Reliable estimates of the incidence of non-paternity are few and far between, although various rates are quoted in an authoritative manner by several sources. … if one attempts to trace the source of such estimates they often appear to be based on hearsay, anecdote or unpublished or unevaluable findings. … non-paternity rates are likely to vary between countries, and to vary within countries by age-group, cultural or ethnic group, region, and age and duration of partnership … It is equally important that the lack of contemporary, verifiable and relevant data is recognized. … We would be interested to hear of any reliable data”. I still don’t buy it. Blood distribution for the UK is: O+ 37.0% A+ 35.0% B+ 8.0% AB+ 3.0% O- 7.0% A- 7.0% B- 2.0% AB- 1.0% The wrong father would need to be of a blood type that contradicts what the expected father is; i.e. AB makes it easy to determine the paternity, but that’s 4% of the population. The only possible fairly common combination would be an couple that is both O and a hidden father not O. A and B could both have O recessive and 50% chance of having an O child. It would be a fun exercise to consider the odds, picking at random, of discovering a paternity error for every threesome combination of blood types. For example, and O and A have a child - which combinations are impossible? B and AB, so 14% of the possible blood types. The child could be A or O, also + or - and unless both parents are - and the child is + there is no contradiction. Ditto for O and B; any parent and AB, only O is impossible. etc. This is why I think the story is BS - to get a 30% “definitely wrong” parentage, you’d need a FAR higher actual parentage error. I don’t think any social class has that level of soap opera going on. ---
**Author:** ![Darren\_Garrison](https://sea3.discourse-cdn.com/straightdope/user_avatar/boards.straightdope.com/darren_garrison/32/92_2.png) [@Darren\_Garrison](https://boards.straightdope.com/u/Darren_Garrison)\ **Post date:** [December 28, 2017, 10:52am UTC](https://boards.straightdope.com/t/what-should-i-conclude-from-these-dna-tests/805242/33 "2017-12-28T10:52:43Z")
Well (playing devil’s advocate here) they _could_ have been looking at [more than just ABO](https://en.wikipedia.org/wiki/Human_blood_group_systems). ---
**Author:** ![KneadToKnow](https://sea3.discourse-cdn.com/straightdope/user_avatar/boards.straightdope.com/kneadtoknow/32/3999_2.png) [@KneadToKnow](https://boards.straightdope.com/u/KneadToKnow)\ **Post date:** [December 28, 2017, 12:23pm UTC](https://boards.straightdope.com/t/what-should-i-conclude-from-these-dna-tests/805242/34 "2017-12-28T12:23:49Z")
_Mazel tov!_ > [@Qadgop\_the\_Mercotan](#): > > I’ve enjoyed my 23andMe experience tremendously. I’ve found DNA relatives whom have helped me take the lineage back farther, and wonder of wonders, discovered I’m 2.1% ashkenazi jew!!! Woo hoo!! ---
**Author:** ![md2000](https://avatars.discourse-cdn.com/v4/letter/m/73ab20/32.png) [@md2000](https://boards.straightdope.com/u/md2000)\ **Post date:** [December 28, 2017, 1:58pm UTC](https://boards.straightdope.com/t/what-should-i-conclude-from-these-dna-tests/805242/35 "2017-12-28T13:58:09Z")
> [@Darren\_Garrison](#): > > Well (playing devil’s advocate here) they _could_ have been looking at [more than just ABO](https://en.wikipedia.org/wiki/Human_blood_group_systems). I guess the next question would be how routine and how expensive would that level of test be in the 1950’s and 1960’s? IIRC a common issue in those days in paternity tests was whether the alleged father could possibly be the father - but more often than not it did not rule out the father; even with AB involved, it was still possible that someone else was the father. then you’re back to the dominant-recessive issue. A dominant gene could be paired with a recessive one; so an A, B, or O mother paired with an A,B, or O father could produce an O, or whatever one of the parents is, A or B. Only an AB father (or mother) cannot produce an O child. Presumably the same applies for many of those other characteristics in the linked article. Without DNA, all you can do is narrow down the odds unless there’s a direct contradiction of dominant characteristics. In the days before DNA tests, it was not simple to determine what recessives were carried except maybe an in-depth family tree study. Which of course brings us back to the premise of the studies… ABO and Rh are pretty common tests because it is necessary to know for hospitalization or blood donation; what are the odds an “unpublished study” would be able to afford hundreds of more in depth tests? ---
**Author:** ![Chronos](https://sea3.discourse-cdn.com/straightdope/user_avatar/boards.straightdope.com/chronos/32/134_2.png) [@Chronos](https://boards.straightdope.com/u/Chronos)\ **Post date:** [December 28, 2017, 2:27pm UTC](https://boards.straightdope.com/t/what-should-i-conclude-from-these-dna-tests/805242/36 "2017-12-28T14:27:45Z")
You usually can’t conclusively say that any given case is a nonpaternity event, just from blood types. But you should be able to calculate the overall nonpaternity rate from the degree of correlation. A father with type A blood won’t always have a child that’s A or AB, but he should have A or AB children more often than the population average. And you can calculate exactly how much more than average, if you know the relative frequencies of all of the types. If, then, you find that it’s less than that, then you can attribute that to a certain percentage of nonpaternity. EDIT: > [@](#): > > Only an AB father (or mother) cannot produce an O child. Likewise, an O parent cannot produce an AB child. And a pair of Os can have only an O child, and a pair of As or an A and an O can’t have a B (and vice-versa). ---
**Author:** ![gkster](https://sea3.discourse-cdn.com/straightdope/user_avatar/boards.straightdope.com/gkster/32/14107_2.png) [@gkster](https://boards.straightdope.com/u/gkster)\ **Post date:** [December 28, 2017, 8:51pm UTC](https://boards.straightdope.com/t/what-should-i-conclude-from-these-dna-tests/805242/37 "2017-12-28T20:51:42Z")
Here’s a chart on paternity and blood types: > **[BLOOD TYPE CHART\_CHILD FATHER MOTHER-PATERNITY determination by blood type -...](https://canadiancrc.com/Paternity_determination_blood_type.aspx)** > > BLOOD TYPE-CHILDREN BLOOD TYPE MATCH TO MOTHER FATHER - PATERNITY FRAUD The chart shows that there’s a lot that can’t be proven–for example, if one parent is type A and the other is type B, their children could be any of the 4 types, including O. ---
**Author:** ![PoppaSan](https://sea3.discourse-cdn.com/straightdope/user_avatar/boards.straightdope.com/poppasan/32/189_2.png) [@PoppaSan](https://boards.straightdope.com/u/PoppaSan)\ **Post date:** [December 28, 2017, 9:39pm UTC](https://boards.straightdope.com/t/what-should-i-conclude-from-these-dna-tests/805242/38 "2017-12-28T21:39:40Z")
> [@gkster](#): > > if one parent is type A and the other is type B Like my parents > [@](#): > > their children could be any of the 4 types, including O. Like my 8 siblings and I. We got almost every combo. [Previous page](https://boards.straightdope.com/t/what-should-i-conclude-from-these-dna-tests/805242.md?page=1)