Shared DNA through both parents

Recently I gave a webinar about choosing a DNA test and breaking down the differences between AncestryDNA, 23andme, and FamilyTreeDNA. When it came to autosomal DNA, I included the fact that 23andme and FamilyTreeDNA provide a chromosome view of how you share DNA with your matches while AncestryDNA does not, giving you just the summary of how much CentiMorgans are shared and along how many segments. For these reasons I do recommend people who test with AncestryDNA to also upload their DNA onto Gedmatch so they can better visualize some of their matches. A question I get concerns why does this matter? I now have an ideal example to share.

I recently got the 23andMe DNA results for my wife’s 86-year-old grandmother Zoila (the only living great-grandparent of our children), so I now have an example of DNA over four generations. Above left are her first four matches: her son, granddaughter, great-granddaughter, and a “Mr. Garcia.”

Garcia is the surname of Zoila’s mother, but is also a very common Spanish surname. I then clicked on shared matches between Zoila and Mr. Garcia:

This result was somewhat surprising. Mr. Garcia shared 2.72% DNA with Zoila over 12 segments, and then 2.71% DNA over 13 segments with Zoila’s son. (The fact that it’s 13 segments instead of 12 is not what is surprising, as one large segment could be reduced to two smaller segments.) What I would consider extremely unlikely is that Zoila would contribute nearly identical shared DNA over numerous segments perfectly to her son.

Because 23andMe allows you to see shared DNA bit by bit, I then compared Mr. Garcia to both Zoila and Jose along each chromosome.

From seeing the chromosomal comparison, it’s now quite clear what’s going on. While several of the shared DNA segments Jose has with Mr. Garcia are the same as those of his mother, there are also several unique segments both of them have. This means that Mr. Garcia would also have shared DNA with Jose’s father! (I can’t have him tested since he is deceased.) I have not found any way that Jose’s parents are related, but given that they are from the same area of Dominican Republic, a “common cousin” is not surprising. If you wonder how you can have more shared DNA with a match than your tested parent, this is one explanation.

This is the value of seeing shared DNA. If you are just told the amount of DNA and along each segments, and then see shared relatives, I might have concluded that Mr. Garcia was a genetic relative just through Jose’s mother, when Mr. Garcia was actually a genetic relative through both of Jose’s parents.

About Christopher C. Child

Chris Child has worked for various departments at NEHGS since 1997 and became a full-time employee in July 2003. He has been a member of NEHGS since the age of eleven. He has written several articles in American Ancestors, The New England Historical and Genealogical Register, and The Mayflower Descendant. He is the co-editor of The Ancestry of Catherine Middleton (NEHGS, 2011), co-author of The Descendants of Judge John Lowell of Newburyport, Massachusetts (Newbury Street Press, 2011) and Ancestors and Descendants of George Rufus and Alice Nelson Pratt (Newbury Street Press, 2013), and author of The Nelson Family of Rowley, Massachusetts (Newbury Street Press, 2014). Chris holds a B.A. in history from Drew University in Madison, New Jersey.

15 thoughts on “Shared DNA through both parents

  1. Dear Chris:
    Thank you for the clear explanation. I have the results of my DNA test from 23andme but I believe, their estimation of time (year) from which the DNA of each population groups proceeds is an incorrect calculation.
    Have you had a chance to study their statistical way of arriving at the year of precedence?
    I don’t understand all the statistics, but based on the known history of our family those dates do not match the ancestry of my ancestors.
    What do you think?
    Hilda

  2. Many thanks for this example. I need some clarification on the first sentence of the next to the last paragraph: “there are also several unique segments both of them have.” Does “both of them” mean Jose and Mr Garcia? And for a neophyte, are these unique segments the orange segments that stand alone, without corresponding purple segments? Thanks Jane

    1. Hi Jane – “both of them” refers to Zoila and Jose in terms of unique segments they have to Mr. Garcia. Zoila has 12 segments (in purple) in common with Mr. Garcia. Of the twelve, she “passed” seven on to her son Jose. Jose has 13 segments (in orange) in common with Mr. Garcia. Of the thirteen, seven were inherited from his mother Zoila, while the other six can be concluded to have been inherited from Jose’s father.

  3. Interesting article. I can relate as I had a somewhat similar experience. I have submitted my DNA, my 90 year old mother’s DNA and my 95 year old paternal aunt’s DNA to AncestryDNA. My father is deceased so I could not submit his DNA. I just received a 5-8th cousin DNA match with a man through AncestryDNA. AncestryDNA gave me a green leaf that shows that this man and I share the same 6th great grandparents, Hannah Ayer and Andrew Mitchell. Great, except that I am descended from Hannah Ayer and Andrew Mitchell through my mother and my mother is NOT a DNA match to this man. (Neither is my paternal aunt.) Obviously, all my DNA from my mother’s side of the family came to me through my mother. So, even though I believe that it is accurate that this man and I are related through Hannah Ayer and Andrew Mitchell this is not where the DNA match is located. In looking at his tree, I can see that he, too, is related to Hannah and Andrew through his mother. But I noticed that his father’s paternal grandparents came from Poland as did my father’s maternal grandfather. I have not been able to make any progress on this side of my family, but I am guessing that this MIGHT be where the DNA connection lies. This has truly been a HUGE brick wall for me. This man knows where in Poland his paternal grandparents were born. I really don’t know if this is the DNA connection, but I intend to investigate further and see what I can find. This man and I share a DNA match with one other person on AncestryDNA and this person has not submitted a tree to Ancestry. As a start, I have contacted this person to see if we can determine the connection.

  4. Thank you for this illustration. Like many, my maternal lines have lots of “cross pollination” of the gene pool which is all old New England stock. I have begun adding two generations further back to what my FamilyTreeDNA results tell me for relationships. This often leads me to the correct generation to look for common relatives.
    Of course I have now found that there are some common ancestors on my mid-western paternal side with a line originating out of New England as well. So far all my known paternal cousins have come from a different line of German descent. I think that I need to take another look at those matches that I assumed were maternal.

  5. Notice that Ancestry (and 23andMe and other) customers should also upload to FamilyTreeDNA to find more matches. It’s free, or only $19 to get all FTDNA’s tools too.

  6. It is also simple math. A small area like the Dominican republic is going to show cousins after a certain amount of generations. The same with Quebec and the limited amount of friends who came there from the early 1600s to the early 1700s

  7. Very interesting. I loved your presentation in Worcester, MA earlier this year on DNA. DNA can be like watching paint dry, but you were so clear with your explanation I was inspired to learn more. Thank you Chris.

    Eleanor Miller Watson

  8. The segments in question that you think Jose shares with Mr. Garcia only are each how many cMs. Any longest segment under the threshold of 10cMs has a diminishing chance of being a false positive or a population match.

    1. Hi Caith –

      Jose and Mr. Garcia have 202 cM in common, most of which is shown in the illustration along 13 segments, 7 at the same place as his mother, 6 not-
      Of the Seven in common with his mother, the lengths are: 13.57, 11.40, 49.68, 10.60, 10.68, 13.88, 31.62
      Of the Six NOT in common with his mother (i.e. through his father), the lengths are: 9.54, 5.96, 5.09, 14.30, 8.95, 16.48

      For the common shown cM I conclude that Jose inherited through his father, this is 58.52 cM along 6 different segments.

  9. Gedmatch is great, right?? So why every time I try it it gives me a kit number and proceeds to reply
    No kit (number) found.. I’m adding results on last time

  10. Hello. I am curious if you can help me understand something in my own family dna. My bio mother has four children, with four different men. My mother and one of my half brothers share a match (random fourth cousin), but neither myself nor my other half brothers share a match with this fourth cousin. If our bio mom is related to this person from my half brothers dads side of the family 1. Why wouldn’t me and my brothers be related to her as well since we are half of our mother and 2. Would that make her and my brothers father related?

  11. Hi Chris,
    Does the natural randomness of dna inheritence effect the conclusion here and particularly in cases where you get similar cM strengths such as this but Jose’s extra might be located on just one or two other segments which Zoila may naturally not have shared with Garcia? Would you be as confident in your conclusion if the number of unique segments from Zoila and from Jose with Garcia was less?

Leave a Reply

This site uses Akismet to reduce spam. Learn how your comment data is processed.