Ancestry Kit Mixup

Every genealogists worst nightmare.  A DNA kit swap.  You unknowingly receive the results from someone else, and that equally in-the-dark unknown person receives yours.  And you’ll never know unless you recognize the signs and take action to see if it’s your bad luck or overactive imagination, or the answer really is a kit swap or lab error of some sort.

I’ve just spent three months unraveling this exact situation that occurred at Ancestry.com.  The person to whom this happened would like to share her story with you.  We are hoping that if something similar ever happens to you, that you’ll be able to recognize the signs and know what steps to take to figure out if this indeed has occurred.

Let me also say that a kit swap or similar lab error is really quite rare, and in most other instances when people believe their kits have been swapped, they haven’t been, although this certainly is not the first time this has happened.  CeCe Moore reported on another Ancestry.com case in 2012.

We’ll call the lady Jane. Jane’s father agreed to have his Y DNA tested some years ago at Ancestry.com.  Jane submitted his DNA for him and noticed that he had no matches to his rather common surname.  She didn’t really think anything of it at the time, other than being disappointed.  His haplogroup was estimated by Ancestry to be R1b.

As time went on, she ordered Ancestry.com’s autosomal test too for her father.  Ancestry sent another sampling kit, and her father is receiving matches to people who, at least according to their trees, share common ancestors with her father.

Last year, Jane decided to transfer her father’s Y DNA to Family Tree DNA. The markers from Ancestry.com were transferred, and Jane still didn’t have any surname matches at Family Tree DNA.

Jane then ordered the Geno2.0 test for her father.  The results were returned with haplogroup I, terminal SNP I-L22, which were at odds with Ancestry’s haplogroup R1b estimate.

About the same time, Jane upgraded her father’s STR markers as well, and the haplogroup project administrator noticed that while Jane’s father’s lower panels, meaning the ones tested at Ancestry matched haplogroup R1b, his upper panels didn’t match R1b subgroups at all.

Obviously something was wrong, very wrong, someplace.  But what, and where?  Jane contacted me and asked if I would help unravel this puzzle.

I checked Jane’s father’s page at Family Tree DNA, and when she transferred his Geno 2.0 results to his FTDNA page, apparently the transfer confused the software at FTDNA because his results reported both I-L22 and R-M269 as positive, which is impossible since I-L22 is in haplogroup I, only, and R-M269 is only found in haplogroup R.

ancestry kit swap ftdna snps

Unfortunately, this only added to the confusion.

At this point, I downloaded the raw data file from the Geno 2.0 test and verified that indeed, M269 was absent and L22 was present.

ancestry kit swap raw data

Family Tree DNA, thankfully, stepped up to the plate and ran a SNP test on Jane’s father’s second vial.  That SNP test also came back as positive for haplogroup I, matching the Geno 2.0 results.

Just to be absolutely positive, Family Tree DNA sent Jane’s father a third vial and tested the same markers that Jane had transferred from Ancestry.  You can see for yourself – the results are very different.  The results are unquestionable.  Either there was a kit swap or a lab error of some sort at Ancestry where the wrong markers were posted for Jane’s father’s results.  He has been tested three times, from separate vials, at Family Tree DNA with all of the results providing evidence that the Ancestry results were in error.

Marker Ancestry FTDNA
DYS438 12 10
DYS391 10 11
DYS392 13 11
DYS426 12 11
DYS439 13 11
DYS445 12 11
GGAAT1B07 10 11
DYS444 11 12
DYS446 13 13
DYS462 11 13
Y-GATA-A10 13 13
DYS437 15 16
DYS441 14 16
DYS458 17 16
DYS463 24 21
DYS635 23 21
DYS452 30 31

In an overabundance of caution, Family Tree DNA is going to rerun the entire test, all markers and the backbone SNP, from yet another (fourth) new vial being sent to Jane’s father.  Thank heavens Jane’s father is still available for testing and not entirely discouraged.

Jane is ecstatic, because now, she is actually receiving surname matches and in her father’s words, “we just wanted to know who we are.”  And just in time for Father’s Day!

Signs and Signals

How might you know if a kit swap has happened to you?  As we know, Ancestry has discontinued their Y and mitochondrial DNA testing and will be destroying the data base, so this won’t be an issue at Ancestry with new Y DNA kits, but it could be an issue for results already delivered, like Jane’s, and for autosomal tests.  This is one reason why retesting might not be a bad idea, even though the $19 or $58 Y DNA Ancestry to FTDNA transfer price is quite attractive.  Here are some of the signs that might tip you that there is a problem:

  1. If Y DNA, you don’t receive any surname matches, even to those you believe that you are in related to. This is one of those sticky-wickets, because if you don’t match your first cousin, for example, the most likely situation is that you have an undocumented adoption in one of the lines. My suggestion in this situation is to submit an entirely new test under a new kit number. If your first and second kits match each other, then the answer is the undocumented adoption.
  2. If autosomal DNA, and you have no matches to anyone you believe you should match, especially close relatives, submit your DNA to one of the other three testing companies – Family Tree DNA, 23andMe or Ancestry.com. The approach gives you the benefit of fishing in multiple ponds along with verifying that your results match each other. When you receive the results from both companies, download the raw data files from both to www.gedmatch.com and then match them to each other. They should match almost exactly, although there will be some small differences in terms of areas tested and possibly no-calls – but they should match very closely.

Let’s hope this never happens to anyone else.  The sad thing is that whoever, at Ancestry, received Jane’s father’s Y DNA results likely has no idea they are incorrect.

Thank you Family Tree DNA for going above and beyond to resolve this very distressing situation for Jane and her father.

______________________________________________________________

Disclosure

I receive a small contribution when you click on some of the links to vendors in my articles. This does NOT increase the price you pay but helps me to keep the lights on and this informational blog free for everyone. Please click on the links in the articles or to the vendors below if you are purchasing products or DNA testing.

Thank you so much.

DNA Purchases and Free Transfers

Genealogy Services

Genealogy Research

34 thoughts on “Ancestry Kit Mixup

  1. That happened to my husband and I found it right away. The company that switched the results did everything to make it right and said that there was a mistake made. No one is perfect and mistakes can happen and it is the way they we handle it that made a difference. They took ownership and made it right.

  2. WOW…that is awful. Errors do happen, but as Ancestry has never done SNP testing for the Y, they have made many errors on haplogroups. BUT, the errors for the marker results! What a mess! Great job Roberta and FTDNA for helping this family!

  3. I was wondering if that means that the people who were tested at Sorenson Molecular Genealogy Foundation will have their DNA destroyed too?

      • I talked to an Ancestry.com representative at the SCGS Jamboree in Burbank, California 6 days ago and they told me that the SMGF DNA samples are not going to be destroyed. A FTDNA customer was incorrectly given my wife’s Family Finder results last year. It took a while, but eventually the customer received his correct results. I know that DNA swaps have occurred at 23andMe at least once previously. Therefore, incorrect autosomal DNA results have been received by customers of all three of the major genetic genealogy companies that do autosomal DNA testing at least once previously.

  4. It just happened to me at FTDNA, only four persons(I believe that an entire batch was misappropriated) results were impacted.

    As a child of two adopted parents, the errors were almost cruel. All three of my labs (Y, Mito, and FF) were misappropriated. The ONLY reason that I caught it was because my Y match and I came up as identical twins on our FF results. After a rigorous line of questioning to our parents (he is from Finland!) by us, it became a point worthy of further discovery. Afterall, the exact match on the Y @67 markers seemed acceptable. According to FF I registered a Finnish component, his father had two paternal surnames- both of which showed up in the next county over from my family in Vermont, so had it not been for FF results matching us, we would have NEVER known.

    Upon contacting FTDNA, they have just concluded re-running my labs with a fresh sample and my Mito labs were wrong too! I went from L0a to H1c!

    It was told to me that an intern had mixed up an entire batch. Honestly, there is very little that can be done in discerning a false report. It was due to my insistence that this was discovered, FTDNA never would have known.

    So now, if I ever receive an exact Y match, I am goung to insist that new labs be run.

    • DNA Enigma: How long ago did this happen/ Several of the tests for my relatives have shown as non matching. I didn’t question this at the time, just chalked it up to being an NPE.

      • The batch mix-up was in February- I now know of one additional kit outside of the three that were impacted in my particular snafu, that was impacted. In all instances, FTDNA is working to get the proper results sorted out- but honestly, this definitely impacts my confidence level in the results.

        I co-admin or admin several projects, one in particular is for L0a (Mito Eve’s haplogroup, for which I was originally assigned to) where there is another man who has no known African ancestry, and has zero African admix in his FF results. I am on the fence about contacting him about the possibility of his results also being incorrect.

    • Family Finder tests are on 24 samples per plate (same as 23andMe now), so “only four persons … entire batch” isn’t consistent. After 23andMe’s one reported mishap, due to a plate having been loaded into scanner wrong way around, the lab made it physically impossible and 23andMe started checking that sex-chromosomes of the results matched the samples. Granted both measures should’ve been in place from the start, and individual errors in labeling/placing samples are always a possibility. It’s always been worrisome to me that FTDNA doesn’t even report sex-chromosomes (although you might be able to check if X-chromosome is single or double), but I guess Y- and mt-DNA present whole another bag of troubles – at least they’re “only” used for genealogy and generally not health. With larger databases to compare again, it’s luckily becoming pretty obvious when this happens.

    • At one time the Geno 1.0 test could be transferred, but I’m not sure if it can be now or not. The major benefit is for the SNPs for the Y test for the Geno 2.0 version.

  5. I sent my DNA to ancestry.com I don’t match any of the names. I have written family history back to 1605 with births deaths marriages etc none of these names show up. I have lots of family out of Canada that didn’t show up. When I called to ask about it was told I don’t match my father or mother, and DNA doesn’t lie! I look like my dad and aunts and brothers and mother. What is going on here?

  6. Pingback: Friday Finds – 06/20/14

  7. Pingback: DNAeXplain Archives – General Information Articles | DNAeXplained – Genetic Genealogy

  8. How often do you think this is happening? I questioned their analysis, because relatives didn’t match and their admixture report was off. But now I suspect there was a mix up in my case too. I recently used some other sites that analyze SNPS and predict traits. Some results didn’t concern me much, because they were experimentalooking or lower confidence predictions. But the good or high confidence ones looked like another person, literally “looked” as in the mirror, not “are you likely to study, be nervous, etc” but physical traits objectively wrong.

  9. I have a situation which makes me think that AncestryDNA has made an error; I spent nearly an hour on the phone today, and only managed to talk to people who do not seem conversant with DNA terminology – they kept repeating patronizing statements that do not apply. I’m looking for suggestions on what to do next.

    My father had autosomal testing at FTDNA. Later I had autosomal testing at AncestryDNA.
    My father had autosomal testing at AncestryDNA. I had my Ancestry testing transferred to FTDNA.

    At FTDNA there is a family group that we match.
    The father matches my father at 46cM; his nephew matches Daddy at 48 cM; both now deceased.
    The daughter matches my father at 55 cM (longest segment 17 cM); predicted 2nd-4th; x-match. All three family members match my father & I at the same location on Chromosome 1.

    The daughter recently did autosomal testing at AncestryDNA, and just received results.

    AncestryDNA states that the daughter is my match, at 6.5cM (!), 5th-8th cousin.
    AncestryDNA does not show the daughter as a match to my father, at all.

    I can understand that different companies do things a little differently, but it appears that they ‘found’ a small match elsewhere (perhaps in common with my mother’s family); and completely ignored the matches on Chromosome 1 (without a Chromo browser, hard to tell).

    What recourse, resources, next step should my match and I consider? (We have been communicating for some time, but our efforts are somewhat hampered by an adoption situation. She took the test at AncestryDNA in order to facilitate discovering our connection, so this is rather frustrating and disappointing.)

    • You don’t really have any recourse. Ancestry uses Timber to strip out segments they feel are too matchy. It’s widely believed in the genetic genealogy community that many of these segments are familiar and not ancient. That may be what you are seeing. The best you can do is to upload elsewhere, at Family Tree DNA and GedMatch and see if the segments match there. I wish there was a better answer.

      The other possibility is that based on when you tested, your transferred matches to FTDNA for anyone who tested after May of 2016 will include only about half of the same SNPS in common with FTDNA participants as people who tested at Ancestry prior to May 2016, when Ancestry changed their chip.

      • Thank you for your response. I will encourage my match to transfer her AncestryDNA results to FTDNA, and see how it compares to her earlier test at FTDNA.

  10. I feel my kit at FTDNA has been mixed up as well. A number of paternal cousins and I tested at ancestry.com. We all match each other as expected. In instances were one doesn’t match all of us, they match the others. Which, from my understanding, isn’t uncommon with autosomal DNA, depending on how far back the connection goes. Anyway, I took those results and uploaded them at FTDNA as well. I also uploaded the ancestry.com data for family members and myself to a Y-DNA prediction site. There were some predictions in which non of us shared, which i discarded. We all have Haplogroup J as a prediction. Which makes sense given the fact our family are Moroccan Arabs. So, next I tested my Y-DNA at FTDNA. It came back as R1b. Which I found to be odd, given the family history and the Y DNA prediction site. At 37 markers, with a Distance of 1 and 2, the majority of the people I matched on FTDNA via YDNA, at 37 markers, have the a English surname. Again puzzling given the fact our family is from North Africa and before that Yemen, and we have a typical Arabic last name. I found all of this very odd.

    Another interesting, though possibly unrelated fact is my batch was supposed to be processed between Feb 12th to Feb 26th. However, I received “my” results at the end of January. Almost a month early. Which again makes me think this is not my kit.

    Some of the above mentioned Y-37 matches also took the Family Finder test on FTDNA. I on the other hand uploaded my results from ancestry.com and paid the extra fee to have my results show up in family finder. NONE of these people match me via autosomal DNA, which I find odd. Again, these are people at Y37 with a distance of 1 and 2. I told this to a rep from the company and I was told that since these people are 10 generations or more back they wont show up via autosomal DNA. Makes no sense to me. How does he know the generational gap between these individuals and myself? If my father takes a Y-37 test, and shows up as a match for me, at Y-37 (because that is the test we both would have taken) does that mean he is 10 generations away from me as well? Y-37 testing doesn’t mean they are ten generations back, just means they only paid for the Y-37 test, but with a distance of 1 and 2 that is extremely close. They should show some relationship via autosomal DNA I would assume.

    I am in the process of having my DNA analyzed via another company, which will also tell me my haplogroup. I have contacted FTDNA and they said they would confirm my results. My problem is, even if they confirm the results, I don’t believe these results are mine to begin with. I have stressed this fact to them a number of times, with little impact. I guess I need to also have male family members Y-DNA tested at FTDNA too, in order to prove or disprove what I am saying. Though I am concerned about their QA, given the number of similar situations I have read on the web.

    • The easiest way to tell if both tests are yours is to see if you match yourself. You said that you took the FF test and you uploaded your Ancestry results. Obviously those are two different kits, since you can’t have two autosomal tests in the same kit. Therefore, you should be your own highest match. If not, then something is clearly wrong. I’ve also known people that uploaded the wrong kit from Ancestry to FTDNA when they manage multiple kits.

      • Ah, maybe I explained incorrectly. I took the autosomal test on ancestry.com. I then uploaded those results to FamilytreeDNA and then paid the 20 dollars or however much it is so that my results would be matched with other people who have tested with FTDNA or who have uploaded their autosomal results from another testing company to FTDNA.

        Im positive I uploaded the right kit though.But the issue isn’t with the autosomal test. The issue is I took the Y-37 test, and it is saying my haplogroup and family member matches are something I am 100% positive they are not. There are people who took the Y-37 test, who supposedly match my kit. These same people also took the FF test. These people are a distance of 1 and 2 at 37 markers. They should match me autosomally as well. Yet none of them do.

        • The people who match you on Y most often don’t match you on Family Finder. The only people guaranteed to match are known second cousins or closer. Only 90% of third cousins match, and from forth cousins more distant, even less.

  11. Hi,
    Can a DNA test results be correct for your Maternal line, but incorrect (someone else’s results) for your Paternal line?
    My results came back in Dec. Zero matches to anyone on my father’s side (but very small family)
    but there were 2 oddities.
    1. My father is Jewish, and I did not show any Jewish markers.
    2. There were two 2nd cousin matches from two unknown families. Luckily both have very large trees built. So far we have not been able to cross reference any family members between those two trees.
    This morning my brother’s DNA results have come back, zero matches to the two unknown trees, and he is listed showing as 1st cousin (aka) 1/2 brother.
    To say I am shook up, that it is showing my Dad is not my Dad would be an understatement.
    Please share your thoughts.
    Thank you,

Leave a Reply to Roberta EstesCancel reply