Last year my enchantment with the series Finding Your Roots with Henry Louis Gates, Jr. inspired me to get tested with 23andMe. Getting those results reminded me of the time when my most nutrition-minded aunt introduced me to Yoplait strawberry banana yogurt back in the late 70’s. After that first taste of creamy tartness, I felt that the best way to honor thats experience was to sample the many, many flavors of yogurt I encountered in the coming years. Just like I once felt compelled to taste boysenberry yogurt, both in blended and fruit-on-the-bottom style, I also tested with AncestryDNA and FamilyTreeDNA.
The marketing for such tests emphasizes their weakest point, geographical analysis of one’s autosomal DNA (chromosomes 1-22, which are inherited from both maternal and paternal ancestors). The variances between my results from these three companies show that precision of admixture estimates is still evolving. All three show that my ancestry is almost entirely European in origin, but they differ sharply on the sub-continental level:
After having tested with all three of these companies, I avoid putting too much importance on the geographical results alone, especially in the trace regions. All three of them found a bit of DNA that they didn’t want to assign to Europe, but they can’t agree on what area of the world most closely matches these segments. I am inclined to think that it is not outlandish that I have a trace of ancestry from Asia, as my mitochrondrial DNA (mtDNA) haplogroup is U2d2, which hails from west Asia. This DNA is passed unchanged from mother to child with rare mutations, so my mother’s-mother’s-mother’s mother (back hundreds of generations) was not European.
The matching component of such tests is much stronger, especially at the second cousin and closer levels. Matching is greatly enriched with the testing of close relatives. I have had the great privilege of having key relations test: my parents, my daughter, my brother, my paternal grandmother (who has since passed away), one aunt, a first cousin, and several second cousins, along with both of my maternal grandfather’s siblings.
With matching, there are only two certainties: a child gets 50% of his/her autosomal DNA from each parent and identical twins are a complete match to each other. I suppose that even these two principles aren’t absolutely certain, for each person has at least a few mutations. Of course, there are also trisomies that would prevent the 50/50 rule, too, but that topic is beyond the scope of this blog entry.
Beyond parent/child and identical twins, there are only averages of shared DNA based on distance between two people in a family tree. If you have siblings, you are not related equally to each one, and there can be significant variance between families in this relationship. For example, my dad and my aunt are 10% more genetically similar than I am to my brother.
Going back a generation, the inheritance of DNA from grandparents can be even more haphazard. Through testing of key relatives, I can tell that I inherited more DNA from both of my grandfathers than I did from my grandmothers. In the case of my paternal side, nearly two out of every three genes I inherited from my dad were from my grandfather. I have matches through him that go back nine generations, which is as far as AncestryDNA’s system will automatically find a tree match.
Before I close this entry, I will briefly discuss two other findings of my testing. First is the revelation that the inheritance of X-DNA is even more unpredictable than that of autosomal DNA. It would stand to reason that a woman should pass on a recombination of her two X chromosomes, but this does not necessarily happen. I passed on the X chromosome I inherited from my mom wholesale to my daughter. My daughter and my mom have the exceptional connection of a complete X match on one of the X chromosomes.
Lastly, do not put too much trust in the rules of blood type inheritance recycled almost as often in soap operas as they are in pop science. Here is the case of my family: Dad O-, Mom O+, Sister O-, Brother O+, and me A+. I definitely have a parent/child 50% match to each of my parents. I read lots of speculation online about people doubting their parentage based on soft evidence while blood type differences are held up as a close second best to DNA matching. My experience tells me that having a so-called “impossible” blood type can sometimes be no more significant than a difference in eye color.