mtDNA Match Discrepancy


At Family Tree DNA, I did the mtFull Sequence test. I have 1,000 matches that include 284 exact matches at GD 0 for my haplogroup K1a1b1a which is a common haplogroup for Ashkenazi.

enter image description here

My Uncle (on my Dad's side) surprises me with 0 matches with his haplogroup H3w. Haplogroup H3w is also common for Ashkenazi.

enter image description here

Can anyone explain this discrepancy?

The only thing I know that may be different is that I initially did the mtFull sequence test. My uncle initially did the mtDNA Plus test and then upgraded to the mtFull sequence test. This shouldn't be the cause of this, should it?


Posted 2017-11-25T23:36:52.980

Reputation: 16 148



I had some comments on this on the ISOGG Facebook group. Ann Turner, James Lick and Rebekah Canada set me straight:

My uncle has 4 mutations from mt haplogroup H3w and no matches that share one or more of those mutations without being different on others to give a Genetic Distance of 3 or less, which is what Family Tree DNA will display.

Whereas, I have 2 mutations from mt haplogroup K1a1b1a but there are a lot of people having the same 2 mutations without having any other, giving a Genetic Distance of 0.

So my uncle is connected to H3w but is on his own. Whereas I am connected to K1a1b1a with hundreds of other.

What this means is that additional branches can and should be added to the mtDNA PhyloTree under K1a1b1a.

Rebekah Canada suggests that I encourage my matches to do the free donation of their mtDNA results to GenBank through Ian Logan's website, so that the tree can be correctly expanded.


Posted 2017-11-25T23:36:52.980

Reputation: 16 148


His mother might have had a heteroplasmy mutation. (That's why I'll never have any matches.)

Or... it just might be bad luck that no one else with his specific subvariant of H3w has tested at FTDNA. (For example, my mt haplogroup is 7 characters long, and my Y haplogroup is 13 characters long.)


Posted 2017-11-25T23:36:52.980

Reputation: 458

How common is that? – lkessler – 2017-11-25T23:43:41.813


For reference, I'm linking to your question:

– lkessler – 2017-11-25T23:46:32.880

I gave two possibilities. Which are you referring to? – RonJohn – 2017-11-25T23:46:38.210

How common is the heteroplasmy? – lkessler – 2017-11-25T23:47:22.233

Out of 274 people in the mtDNA project I belong to at FTDNA, about 20 people have heteroplasmy mutations. Mine specifically is a back-mutation, and there are 5 of those. – RonJohn – 2017-11-26T00:00:32.810

My uncle does have exact HVR1 and exact HVR2 matches and some people who are exact on both HVR1 and HVR2. Lots of them are haplogroup H3w. He just doesn't have any mt Coding Region matches. Is that your situation as well? – lkessler – 2017-11-26T00:05:20.823

No. I have zero matches even at the HVR1 level. Maybe your uncle just has the bad luck of no other close matches having tested at FTDNA. – RonJohn – 2017-11-26T00:17:35.360

Sincerely doubt it's bad luck. There's lots of H3w's tested that match him at FTDNA on HVR1 and HVR2. And he has 14,500 autosomal matches. Thanks for your ideas. I'm starting to wonder if it's a FamilyTreeDNA problem that they can fix. – lkessler – 2017-11-26T01:14:09.483

I added another personal example to the answer. – RonJohn – 2017-11-26T01:39:17.180