Monday, May 11, 2015

mtDNA doesn't change fast enough to be genealogically useful -- or does it?



I just love the scientific method of family history research. I can make an educated guess as to what a family relationship may be. Then I can formulate a hypothesis that is a reasonable explanation for that guess. Next I can set this hypothesis up as a figurative piñata. I then take a figurative stick and try to beat that hypothesis into submission using every genealogical tool I can lay my hands on. If I fail, I invite my family members and other genealogical colleagues to take whacks at my piñata hypothesis. As long as none of us can break my hypothesis, I assume it is correct and move on to formulate additional hypotheses to test. Of course sometimes there are subsequent developments that rattle my old undisproven hypotheses. That's when I have the opportunity to learn something new. Learning something new is one of my favorite things.

According to the conventional wisdom among genetic genealogists, mitochondrial (mtDNA) changes so slowly that it is not often genealogically useful in sorting out relationships. This was easy for me to believe when I went for years without getting an exact match.

Then in my views began to change. In two cases it appeared that mtDNA provide important guidance even though in neither instance are they likely to lead to exact matches with ancestral relatives. These cases have been previously reported in this blog. One related to my paternal grandmother's mtDNA and the other my maternal grandmother's mtDNA.

Those of us who know Judy Russell (AKA The Legal Genealogist) find her to be unusual -- if not unique. Therefore, I didn't pay the proper amount of attention when she posted about her mtDNA mismatches with a first cousin. Then I discovered mtDNA mismatches with a cousin that caused me to write Judy for confirmation. Here is her reply which I repost with her permission:
Yep, and it's a very close relationship: my own first cousin. This particular cousin is the younger of two daughters of my mother's youngest sister. Full sibling situation, no chance of any NPE of any kind, all the right autosomal indications.
And this maternal first cousin of mine and I show as a genetic distance of two.
 It turns out that I have a heteroplasmy in HVR1, and Paula has one in the coding region. FTDNA reports that as a genetic distance of two. Interestingly, her mother doesn't have either of those heteroplasmies, so -- as Paula puts it -- we're both mutants!
Wrote this up on the blog: Getting the drift.
In my case a Sherry, 3rd cousin, and I discovered that we were listed by FTDNA as mismatches of THREE. Are we both mutants? We are still trying to work that one out.



Sherry and I have each identified a first cousin to test to narrow our gap. In so doing we hope to be able to follow our mtDNA two generations back to our respective maternal grandmothers. This would be along the red lines shown connecting the individuals in the above diagram. After we get these test results back we will reassess and consider our next step(s).

In our case we are fortunate that Sherry and I are exact matches in the coding region -- the largest part of our mitochondria. This means we can test additional cousins for a relatively reasonable cost. We only need to order HVR1 & HVR2 tests to help resolve our mutant markers.

The HVR1 location at which we mismatch is 16519 Sherry's value is T and mine is C. The HVR2 location is 152 where I have a C and Sherry a T. I have taken the BIG Y test. Until recently the BAM files that were part of the output for the BIG Y had data on mtDNA. In the analysis of my BAM file conducted by yFull, I found the below report of the findings at two locations in question: 

Search in BAM file
ChrM position:
152 (+strand)
Reads:
9
Position data:
9C
Weight for C:
1.0
Probability of error:
0.0 (0<->1)
Sample allele:
C
RSRS allele:
C
rCRS allele:
T
  

Search in BAM file
ChrM position:
16519 (+strand)
Reads:
21
Position data:
5T 16C
Weight for T:
0.226751592357
Weight for C:
0.773248407643
Probability of error:
0.320675177201 (0<->1)
Sample allele:
C
RSRS allele:
C
rCRS allele:
T

Of particular interest is that there were 21 reads at location 16519. The results were 5 "T"s and 16 "C"s. All 9 reads at location 152 found "C".  

To be continued as reports on cousins Jim and Dell come back from the lab in about a month. As they do will it cause me to have to re-examine my hypothesis about my one exact full mitochondrial match (see above)? Will I find that the actual mtDNA haplotype of my maternal 2nd great-grandmother really was not the same as mine? Stay tuned.


3 comments:

  1. Nice to know that this logic works. I am wrestling right now with one of my lines where the tester has no matches of zero but thirty-some matches at one genetic distance. (Those thirty are zero to one another.)

    I figure we broke off from them. I am considering trying one of the grandchildren of the sisters of the tester's mother to see if maybe the tester's mother herself is the place this originated.

    ReplyDelete
  2. Israel,
    Thanks for sharing. Good luck in your research. The logic is there and it will be nice to see how this plays out within real families. As you are well aware, you face the considerable additional problem of endogamy in much of your research. My daughter-in-law's father (Ashkenazi) has 529 full mtDNA matches (within 3) -- 147 of them exact. At this point his results become a very blunt instrument but possibly useful for grouping extended family.

    ReplyDelete
  3. Many people conflate the terms "useful" and "difficult". For hypothesis-driven experimentation, like cases where ones grandparent may have been adopted, the analysis is useful and clear, and the mutation rate doesn't really matter unless your two control groups are unlucky enough to be the same common type.

    For fishing it is very difficult because the surname changes every generation. And if you have tons of matches, it is going to be difficult to sift through.

    But these difficulties don't have any bearing on the USEFULNESS of the mtDNA strategy.

    ReplyDelete