Over the last year or so I have had some interesting matches amongst “DNA Relatives” on the website 23andme.com. I manage the profiles of my parents, my mother’s brother, and my wife’s parents. So far, the most interesting results have all come about from my mother and uncle. While all the profiles have given me several DNA relatives predicted to be anywhere from third to fifth cousins, only my mother and uncle’s matches have been linked me with cousins with whom we have been able to determine the exact degree of the relationship.
A common question raised when I speak on DNA involves how long we carry the DNA of a given ancestor. While our Y-DNA and mtDNA have specific gender-linked patterns of inheritance, our autosomal DNA is more complex. Many of our DNA relatives are predicted to be our relatives within five to seven generations, but if you have identified a relative within that time frame, and you show no DNA in common, are you not related? Has the DNA linking the two of you faded away?
A recent example in my own family involved identifying two relatives, both of them descended from the Shake family of Jefferson County, Kentucky. Alison is a great-great-great-great-granddaughter of William and Sarah (Smith) Shake, my mother and uncle’s great-great-great-grandparents, thus making Alison and my mother fourth cousins once removed. Norvan is a great-great-great-grandson of Christopher and Elizabeth (Davis) Shake, the parents of our William, thus making Norvan and my mother also fourth cousins once removed, but in the senior generation (i.e. Alison is the daughter of my mother’s fourth cousin, while Norvan is the fourth cousin of my mother’s mother). So they are descendants of the Shake family at three different generations, with Norvan splitting off at one earlier generation (see above chart, males in blue and females in red).
23andme allows you to compare your DNA bit by bit. The amount of DNA shared between us all is as follows (my mother’s name is Joy and her brother is Paul):
While my uncle shares two DNA segments with the two Shake relatives, and my mother only shares one, her segment is one of the two segments that my uncle shares. The amount of DNA they all share is predictable for relatives of this degree of kinship, and the fact that my uncle has just a little bit more “Shake DNA” than my mother is not surprising. What is interesting is when I compare Alison to Norvan, their kinship between each other is fourth cousins twice removed. They have no DNA in common:
Yet they are related through the same ancestors my mother and uncle are, for which a portion of “Shake DNA” was shared. My mother and uncle are one generation closer in kinship to both Alison and Norvan (but in different directions) than Alison and Norvan are to each other, and in this scenario, this extra generation is where their common “Shake DNA” was perhaps “shaken off!”
8 thoughts on ““Shake it off!””
Our extended family has also reached out to 23ndMe and other DNA relative sites. In our family’s past there was an enexpected “encouter” with a relative so that the “paper record of the family tree” did not match the DNA family tree connections. In our case it was between other relatives so that we could connect the DNA “dots” to lead to that conclusion.
With your detailed comparison of DNA segments over various generations, does this “Shaken off” relative’s children have the Shake DNA reappear once again in the next generation?
I have many 23andME DNA relatives where the connection seems impossible to find. Every once in a while a DNA relative knows their detailed distant family tree and all of the dots connect. Our challenge as researchers is to find the connections with the exponientially large number of possible relatives especially when the connections is rated as 3rd to distant cousins.
Bill – From my understanding, a “DNA segment” can be inherited by the child of someone, but if it was not inherited by the child, it would not reappear in a subsequent generation. In this scenario above, there were actual four “Shake DNA” segments. My uncle had all four, my mother had two, and Alison and Norvan each only had one, for which those were different from each other. This isn’t to say this is the entire genetic legacy each of us has from our Shake ancestors, but it’s the only common DNA that has been inherited by us this many generations down. My own results are currently being analyzed, so it would be possible to have inherited portions of the two “Shake DNA” segments my mother has, but it would not be possible to inherited the other two “Shake DNA” segments that my uncle had that my mother did not.
A slight correction to what I wrote in this comment above, Alison and Norvan each had two “Shake DNA” segments (not one)
That Alison & Norvan have no DNA in common is not at all surprising given that you get 1/2 your genes from your mother and half from your father. Who gets what in each generation is random which is why even brothers & sisters can be very different. Granted there have been many changes in the field of genetics since I studied it in grad school, but this fascination with autosomal DNA testing is ridiculous.
With all due respect, I don’t think it is completely ridiculous — as long as you have a specific objective in mind — otherwise it is a lot of data to manage or get lost in. In my case, despite enormous efforts on the ground, we had been unable to determine if our nominal paternal great-grandfather was really our biological ancestor or even if our paternal great-grandmother was really the mother of my grandfather. DNA testing — both “Y” and autosomal — answered those questions which plagued two generations of family historians. And I have learned a lot from tests on other family members. What I am unsure of is just how completely random the process is. It seems from my results — I manage 10 family members’ tests — that there is “stickiness” on some reasonably long DNA segments that survive intact from one generation to the next to the next and so on — throwing off predictions of closeness considerably — the result being MRCA’s who are much further back than predicted. But maybe I just don’t understand probability.
Gentlemen – I’m nearly completely in-the-dark relative to DNA genealogical matching; is it possible that the discontinuities are the result of the “milk man”?
NEHGS’ weekly email carried the article about my cousin Heather as a Black Market baby sold in Butte MT in 1946 and my efforts to find her biological parents..The article has shown up in several newspapers in the west…I did locate two ‘brothers’ in Utah on her maternal side which is where she and I are 3rd cousins…The lingering question for me – a novice in DNA – is the anc.com autosomal just said ‘close’…she then submitted hers and her brother’s DNA to a company that does sibling tests but with only 15 sites…they reported only a .03 difference between half and full sibs…yet another reported half siblings…other evidence makes the half relationship far more likely…So I see two possible explanations: 1] her bio father may have been related to her half-brother’s father [not uncommon in Utah due to intermarriage of same families] or 2] the variance in DNA inheritance means her brother received more of his mother’s than his father’s…is either of these more likely or is there a third possibility I’ve overlooked?