“Why do vendors give me different haplogroups?”
This questions often comes up when people test with different vendors and receive different haplogroup results for both Y and mitochondrial DNA.
If you need a quick refresher on who carries which types of DNA, read 4 Kinds of DNA for Genetic Genealogy.
You’re the same person, right, so why would you receive different answers from different testing companies, and which answer is actually right?
The answer is pretty straightforward, conceptually – having to do with how vendors test and interpret your DNA.
Different companies test different pieces of your DNA, depending on:
- The type of chip the company is using for testing
- The way they have programmed the chip
- The version of the reference “tree” they are using to assign haplogroups
- The level they have decided to report
Therefore, their haplogroups reported may vary, and some may be more exact than others. Occasionally, a vendor outside the major testers is simply wrong.
Not All Tests are Created Equal
All haplogroups carry interesting information and can be at least somewhat genealogically useful. For example, haplogroups alone can tell you if your direct line DNA (paternal or matrilineal) is probably European, Asian, African or Native American. Note the word probably. This too may be subject to interpretation.
A basic haplogroup can rule out a genealogical match through a specific branch, but can’t confirm a genealogical match. You need to compare specific DNA locations not provided with haplogroup testing alone for genealogical matching. Plus you’ll need to add genealogical records where possible.
Let’s look at two examples.
Mitochondrial DNA
Your mitochondrial DNA is inherited from your mother’s direct line, on up you tree until you run out of mothers. So, you, your mother, her mother, her mother…etc.
The red circles show the mitochondrial lineage in the pedigree chart, below.
If your mitochondrial haplogroup is H1a, for example, then your base haplogroup is “H”, the first branch is “1” and the next smaller branch is “a.”
Therefore, if you don’t match at H, your base haplogroup, you aren’t a possible match on that genealogical line. In other words, if you are H1a, or H plus anything, you can’t match on the direct matrilineal line of someone who is J1a, or J plus anything. H and J are different base haplogroups who haven’t shared a common ancestor in tens of thousands of years.
You can, however, potentially be related on any other line – just not on this specific line.
If your haplogroup does match, even exactly, that doesn’t mean you are related in a genealogically relevant timeframe. It means you share an ancestor, but that common ancestor may be back hundreds, thousands or even tens of thousands of years.
The further downstream, the younger the branches. “H” is the oldest, then “1,” then “a” is the youngest.
Some companies might just test the locations for H, some for H1 and some for H1a. Of course, there are even more haplogroups, like H1a2a. New, more refined haplogroups are discovered with each new version of the mitochondrial reference tree.
The only company that tests your haplogroup all the way to the end, meaning the most refined test possible to give you your complete haplogroup and all mutations, is Family Tree DNA with their mtFull Sequence test.
A quick comparison of my mitochondrial DNA at the following three vendors shows the following:
| 23andMe | Living DNA | Family Tree DNA Full Seqence |
| J1c2 | J1c | J1c2f |
With Family Tree DNA’s full sequence test, you’ll receive your full haplogroup along with matching to other people who have taken mitochondrial DNA tests. They are the only vendor to offer Y and mitochondrial matching, because they are the only vendor that tests at that level.
Y DNA
Y DNA operates on the same principle. Specific locations called SNPs are tested by companies like 23andMe and Living DNA to provide customers with a branch level haplogroup. You don’t receive matching with these types of tests.
Just like with mitochondrial DNA, a basic branch level test can eliminate a match on the direct paternal (surname) branch but can’t confirm the genealogical match.
If your haplogroup branch is E-M2 and someone else’s is R-M269, you can’t share a common paternal ancestor because your base haplogroups don’t match, meaning E and R.
You can share an ancestor on any other line, just not on the direct Y line.
The blue squares show the Y DNA lineage on the pedigree chart below.
Family Tree DNA predicts your haplogroup for free if you take the 37, 67 or 111 marker Y-DNA STR test, but if you take the Big Y-500, your Y chromosome is completely tested and your haplogroup defined to the most refined level possible (often called your terminal SNP) – including mutations that may exist in only very few people. You also receive matching to other testers (with any Y test) which can be very genealogically relevant, plus bonus Y STR markers with the Y-500.
OK, But Why Do Different Companies Give Me Different Haplogroup Results?
Great question.
For this example, let’s say your haplogroup is H1a2a.
Let’s say that Company 1 uses a chip that they’ve programmed to test to the H1a level of haplogroup H1a2a.
Let’s say that Company 2 uses a chip that they’ve programmed to test to the H1 level of haplogroup H1a2a.
Let’s say that you take the full sequence test with Family Tree DNA and they fully test all 15,659 locations of your mitochondria and determine that you are H1a2a.
Company 1 will report your mitochondrial haplogroup as H1a, Company 2 as H1 and Family Tree DNA as H1a2a.
With mitochondrial DNA, you can at least see some consist pathway in naming practices, meaning H, H1, H1a, etc., so you can tell that you’re on the same branch.
With Y DNA, the only consistent part is the base haplogroup.
With Y DNA, let’s say that Company 1 programs their chip to test for specific SNP locations, and they return a Y DNA haplogroup of R-L21.
Company 2 programs their chip to test for fewer or different locations and they return a Y DNA haplogroup of R-M269.
You purchase a Big Y-500 test at Family Tree DNA, and they return your haplogroup as R-CTS3386.
All three haplogroups can be correct, as far as they go. It’s just that they don’t test the same distance down the Y chromosome tree.
R-M269, R-L21 and R-CTS3386 are all increasingly smaller branches on the Y haplotree.
Furthermore, for both Y and mitochondrial DNA, there is always a remote possibility that a critical location won’t be able to be read in your DNA sample that might affect your haplogroup.
Obtaining Your Haplogroup
I strongly encourage people to test with and upload to only well-known major companies or organizations. Some companies provide haplogroup information that is simply wrong.
Companies that I am comfortable with relative to haplogroups include:
- Family Tree DNA
- 23andMe
- Living DNA
Neither MyHeritage nor Ancestry provide Y or mitochondrial haplogroups.
The chart below shows the various vendor offerings, including Y and mitochondrial DNA matching.
| Company | Offerings | Matching |
| Family Tree DNA – Y DNA | Y haplogroup is estimated with STR test. Haplogroup provided to most refined level possible with Big Y-500 test. Individual SNP tests also available. | Yes |
| Family Tree DNA – mitochondrial | At least base haplogroup provided with mtPlus test, plus more if possible, but full haplogroup plus additional mutations provided with mtFull Sequence test. | Yes |
| Genographic Project (obsolete in 2019) | More than base haplogroup for both Y and mitochondrial, but not full haplogroup on either. | No |
| 23andMe | More than base haplogroup for both Y and mitochondrial, but not full haplogroup on either. | No |
| Living DNA | More than base haplogroup for both Y and mitochondrial, but not full haplogroup on either. | No |
Want More Detail?
If you’d like to read a more detailed answer about how haplogroups are determined, take a look at the article, Haplogroup Comparisons Between Family Tree DNA and 23andMe.
______________________________________________________________
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
- Family Tree DNA
- MyHeritage DNA only
- MyHeritage DNA plus Health
- MyHeritage FREE DNA file upload
- AncestryDNA
- 23andMe Ancestry
- 23andMe Ancestry Plus Health
- LivingDNA
Genealogy Services
Genealogy Research
- Legacy Tree Genealogists for genealogy research
Share this:
Discover more from DNAeXplained - Genetic Genealogy
Subscribe to get the latest posts sent to your email.
Roberta, this is one of best laid out, clearly understandable articles I’ve ever read on this topic. You make it so clear what one should expect (or not expect) from DNA testing. It is the URL I am sending to all the misguided genealogists who contact me about their DNA tests. Thank you so much.
This is the first time I was able to follow the explanation of what all the letters and numbers mean!
Great article, Roberta! I’m helping an adoptee that had 0 matches at 12, 25, 37, and 67. Is that rare? He’s a J-M172.
I still see no matches quite often.
I’m am so happy to see a report on Family Tree . I still don’t understand my results for my mitochondrial test, it’s a bit much for my brain, but I finally feel relaxed about which company I used thanks to your article. I can also now go to the links you provided for more information and hopefully understand more, so thank you for this article.
Roberta, great article and guide for anyone who wants to learn more about their test results and how it all comes together. Thank you!
Roberta. Thank you for the explanations about this complex world of Dna. I appreciate your bluntness and know that I can rely on your explanations and ideas, This has really helped my understanding.
I’m back to mystified! A male cousin has done the Y DNA and is I-M253. There are a lot of I-M253 Wards on the project. I was looking at another Y project for a different surname and they have plenty of I-M253 people, too. I suppose if I looked at every project I would find surnames with that haplogroup? Is there anything I can do with this information? I’m trying to get a second male to take the test but all are afraid the government will find out something about them. With as many cousins who have tested (autosomal) I doubt the government will find anything interesting about them specifically if they would test!
Look at who they actually match at FTDNA. The haplogroup isn’t useful except to exclude someone.
Question: If two people have a genetic distance of say 2 on 111 STR markers but their haplogroups are reported differently e.g. RS5668 and RZ16337 then could they share a MRCA within a few generations? If the answer is yes, is it because RZ16337 is at a lower level to RS5668 under L513? Or is it way more complex than that?
Look at the haplotree on their account at Family Tree DNA. You can see what is above what.
Thank you for this article. I did a Y-DNA test for my brother on Family Tree DNA. He has pages of matches but none with our family surname. He also has more than a dozen matches with a genetic distance of 0. All but two have different surnames than the others. Is this common? I know who my paternal great grandfather is (and I have matches on ancestry of known cousins to conform it) I’m trying to figure out how I can use the Y-DNA information to trace this line further back. Any ideas?
It depends on whether your brother took a 12, 25, 37,67, or 111 marker test. On a 12 marker test you will match many people of different surnames. As you take a more advanced test such as the 25 or higher, the number of matches with persons with a different surname will decrease with each higher marker test, and your match will be with those persons who share your surname.
My situation is a bit more…complicated. I know that my Paternal side came from Asia, from Korea, since my Great-Grandfather immigrated here in 1905. I am a descendant by barely two generations after his time now. I did a DNA test by both 23andme and FTDNA, both roughly told me the same thing. That my Y DNA was O-M1204 (47z), which is unusual enough for Asian standards, also because he was from what is now the DPRK or ‘North Korea’. I just took the DNA test with Helix or Geno 2.0 and their Y DNA result is dramatically different. It says that I am ‘R-L144’ (U-152), which is European, and an equally rare one at that. I know of my European relatives on my Dad’s side that this could draw from, but it makes little sense obviously. I wrote them, sent them my data and info about my Korean relative, etc, but nothing has changed. So…what do you think is going on?
My mtDNA is always correct generally from everywhere however, it is K1a1(c) or so.
You did the right thing to contact them. I’m surprised at the result. Perhaps you had a read error in a particularly bad location that is used for haplogroup determination.
You should have them recheck your test. We recently had a man take the Ancestry,com y-DNA test. His test results were not a march with any surname. When he had his Ancestry results transferred to FTDNA, an error was discovered with several of his numbers. He had an Allen surname but was shocked to learn that he still did not match any Allen Surnames, but was instead a 67/67 match with more than a dozen members of our Clements Surname Group. The important part of my reply is that a mistake was made in the lab report on his first test. If he had not requested a transfer of his results to our lab, the error might never have been discovered.
Ancestry.com hasn’t done Y DNA testing in many years and there hasn’t been matching since they obsoleted that product. Some of the Ancestry numbers are counted differently than FTDNA numbers. Glad he transferred.
This article was helpful and clear! Thankyou. my question is I used CGI and my haplogroup came back different from my mothers. Because she has dementia I cant even ask questions. How can this happen? Does this mean I am not her natural daughter? How do I find out for sure? I am So distressed. I called the co. and got 3 different answers from 3 different people. now what?
Did you both test with them? CRI is not recommended. I would suggest you both test at Family Tree DNA. It’s a swab test so you could probably help her do it.
Many thanks first for the clarifications.
Could a prediction of a haplogroup be erroneous before snps tests because of what we call STR haplotypes CONVERGENCE
Thank you
It’s theoretically possible. But the pattern for the base haplogroup is pretty distinctive. That’s why only the high level haplogroup is predicted.
Thanks again
I know that’s factually possible based on my knowledge of someone who has been assigned a distant predicted haplogroup from his “natural” father’s(confirmed)
The testing company is of a great fame, thus he’s still waiting for the BigY results and the signs(+/-) that would be given to the snps under analysis.
It seems that there’s a “tricky” convergence between E and I2.
That’s exactly why we need SNP testing.
Hi Roberta,
Good explanation, thank you!
I have two question and I would be glad if you could point out the right direction:
1, I would like to understand how different haplogroups can exist if everybody descend from a common ancestor?
2, Is there any haplogroup match (mitochondrial and/or Y chromosome) between human and primates that shows our relations?
Mutations happen and are inherited by descendants. Those mutations are the branch points for new haplogroups. I have a series of help articles for both Y and mtDNA if you look at the top of the main page. Also, the relationship between humans and primates is very ancient. Yes, scientists do use that data.
I have an uncle who tested Y as: R-M269 (which would be the correct halpogroup for the line) then his brother tested and he was a Q halpogroup on FTDNA. BOTH are related to me through a male relative meaning they can only be related to me through their father the on the Y: R-M269.
They are DNA matching me in the correct amounts I am not sure why is this happening? The site tried to say oh well we can just test you again. That was not the explanation I needed considering the huge failure on the first go round. It seems to me the Q was contaminated sample or the chip is faulty. Are there actual real labs that can run this not using the FTDNA chip? Their chip or method seems faulty in our case.
I think maybe using an outside lab might help us any suggestions would be great. 😀
The only way to find out if it’s contaminated is to rerun it. They are an exceptional lab. You didn’t mention if you got the haplogroup via autosomal or actually took the Y-DNA tests for all men involved. Do they match each other? Also, did you all test at their lab or was someone’s DNA uploaded. I’ve seen the wrong files get uploaded and assigned to the wrong person.
Yes they match each other they are DNA full brothers. 1 uncle uploaded from from “Ancestry,” the other tested at their lab. My uncle who uploaded from “Ancestry” had no other samples only his own so this was not a mix up with the tests. R-M269 would be correct for our ancestor and for other males of our line. Our family is very well documented and researched by anthropologists, genealogists and DNA researchers as they are well known founders. The founder of the line is R-M269 from Northern Spain.
When I look at the surname study on FTDNA and the results also for this surname there seems to be slight mutations in string numbers but more or less identical to most of the surname matches with R-M269.
There is nothing that can explain an entire change of a halpogroup even looking at the string. It seems like the chip is seeing R-M269 uploads as Q! I think the problem is with the chip. Would you know how this chip is being calculated?
I know supposedly: Haplogroup Q is thought to have originated in Central Asia or North Asia during of shortly after the Last Glacial Maximum (LGM, 26,000 to 19,000 years ago). Q descends from haplogroup P, which is also the ancestor of haplogroups R1a and R1b. I think the chip is basically viewing things that descend from P as either Q or R. That is why it is important to know how their chip is making this calculation.
I believe my uncle did redo the test using their test and now it is citing Q but only after he informed them that his full brother is a different haplogroup which is impossible. So the problem seems to be with the chip. The numbers have not changed in the string they match the R-M269 surname identified by DNA genealogists. I have never seen anything like this. I think we will have to go to another lab. If you know of another lab to suggest I would greatly appreciate it.
Thanks again Roberta!
You need to call support at FamilyTreeDNA.
I did manage to call them they didn’t understand what happened either they checked to see if he had uploaded from Ancestry they didn’t find that. Evidently he took the Y-37 and got R-M269 then he talked to them about why his full DNA brother’s Haplogroup was different and suggested he take the Big -Y and that resulted in a Q assignment. The Q is most definitely not correct. An R does not become a Q no matter how defined they test the DNA.
I think their chip must be the one that has the trouble. Q simply is not possible given everyone in relation to each other and they are the only ones with Q now with everyone else R-M269. So this isn’t a matter of different fathers as again they are full DNA brothers. FTDNA were equally baffled. They suggested unloading ancestry again to see if the result is different to the FTDNA Q designation. We are VERY disappointed about this process so wish us luck. 🙂
I am willing to follow up with you on this, but I can’t without the kit numbers of the people involved. You can email me privately if you wish. I have never seen this issue – not once, with any FTDNA result. I have seen the issue with uploaded files because one of the necessary SNPs is a no-call which forces the haplogroup identification further up the tree.
Hi,
Here’s my question: my fathers’ dna is R1a, etc. and I’ve tried to see if my cousins have the same R haplogroup. They don’t–they have R1b. How can that be?
Be sure that you are comparing results at the same vendor from the same test. They are normally not called R1a or R1b anymore so it could be an out dated source. Also, be sure you’re comparing the direct paternal lineage. The actual answer is that it can’t be but you need to insure consistency and accuracy.