How to tell the relationship from the shared DNA

Many people have the illusion that if their testing company says a person is a 3rd to 5th cousin they really will be. That is not the case.

The testing companies are just making the best guess they can from the data they have. They do not seem to take segment sizes into account, rather they primarily use total shared DNA measured in centimorgans (cMs) for their relatedness estimates, usually the sum of all matching segments of 5 cM or larger. Close relatives will always share larger chunks with each other and so size does matter here.

Recently I have received numerous questions from people trying to figure out if a new match is a half sibling or a niece or a grandchild. These are hard to tell apart without testing more relatives as they all share about 25% of their DNA with each other. So I decided to collect some detailed statistics on those specific relationships with a google form (click here) that includes total segments and segment sizes for a future blog post.

The companies predict reasonably well for close family but it is just not possible to be accurate beyond that due to the randomness of DNA inheritance.

For example, here is a picture from the new 23andme of some of the DNA I share with Dick, a 2nd cousin on Dad’s paternal side so blue, and John, a 2nd cousin on Dad’s maternal side so red.

Ktty2ndCousinsS

I share a third again as much DNA with John as I do with Dick,  even excluding the 14 cM on the X. The expected amount for a 2nd cousin is 3.125% which is 212.50 cM, right in the middle between these two.
Henry2nds Checking my brother, I see the same effect – he has 282 cM with John versus 185 with Dick. Not surprisingly, when I look at Dad I find that he shares almost twice as much with John as with Dick. Clearly he just inherited more of the same DNA as John’s mother from their common grandparents. Conversely, he inherited less DNA shared with Dick’s mother from his other grandparents.

On the left is a comparison of my first cousin Henry with both Dick and John. The amount he shares with each 2nd cousin is practically identical, as long as you subtract the 40 cM that he shares with John on the X from the total shown by 23andme. Amazing how variable DNA inheritance can be among 2nd cousins.

Click here for the ISOGG wiki article on Autosomal DNA statistics which usually includes the current chart from Blaine Bettinger’s shared centimorgan DNA project.

Personally I have his chart (shown below, click it for a larger version) bookmarked for easy reference. I rely on it heavily.

SharedcMProjectSmll

Warning, Ancestry.com DNA testing will show a smaller number of matching cMs and larger number of segments due to their algorithm which removes population specific segments.

The DNA adoption site has a relationship calculator that can help figure out closer relationships discussed in the article at Roberta’s blog called Demystifying Ancestry’s Relationship Predictions Inspires New Relationship Estimator Tool.

Autosomal DNA matching is not cut and dried due to the randomness of DNA inheritance and is even more confusing if you are from an endogamous population because your parents will likely share some DNA due to ancestral cousin marriages. Thus a match could be related on both sides! There is a function on the GEDmatch site that lets you check if the parents of a specific kit are related because they have passed along matching DNA segments.

I have sometimes found that someone predicted to be a 3rd/4th cousin based on total cMs is much more distant. This has happened when there are two good sized matching segments but each segment is from a different ancestral couple. Thus the relationship is much further back, for example, a double 6th cousin.

Another issue is the fact that the testing companies cannot tell which of the two paired chromosomes a match is on. So when you have a match that neither parent has, it is a false match created from small bits from each parent by the computer program (see my IBC article). This is why I prefer to look at matches that are “phased” that is to say a child and a parent have the same match.

If you only match someone on a single good sized segment (greater than 10 cM for most, more than 20 cM for the endogamous) your DNA relative can be anywhere from a 4th to a 14th cousin. See http://ongenetics.blogspot.com/2011/02/genetic-genealogy-and-single-segment.html?m=1 for a further discussion of that.

Last but not least here is my data collection form which you can fill out right from this blog post (use the slider on the right to scroll down it to answer all questions).

15 thoughts on “How to tell the relationship from the shared DNA

  1. Thank you for this! 🙂 Especially with the half-sibling vs. grandparent to grandchild or aunt/uncle to nephew/niece relationship, which are all difficult to tell.

  2. In my project using 23andme comparisons there is a difference in segment counts for one relationship vs another.

    Grandparent-grandchild from 23 segments to 37 segments. The 37 segments is an outlier with 3 X-chromosome segments. Grandmothers appear to share a couple of more

    Ethan vs. Robert, 21.9%, 1629 cM, 23 seg (zero X-chromosome)
    Ethan vs. Janice, 25.4%, 1892 cM, 26 seg (zero X-chromosome)
    Ethan vs. Rocky, 22.0%, 1638 cM, 24 seg (79cM X-chromosome, 2 seg)
    Ethan vs. Vivian, 28.0%, 2082 cM, 37 seg (103cM X-chromosome, 3 seg)

    Kalea vs. Robert, 24.1%, 1792 cM, 25 seg (zero X-chromosome)
    Kalea vs. Janice, 25.7%, 1909 cM, 26 seg (full X-chromosome, 182cM)
    Kalea vs. Rocky, 28.1%, 2089 cM, 23 seg (full X-chromosome, 182cM)
    Kalea vs. Vivian, 21.9%, 1631 cM, 25 seg (zero X-chromosome)

    Shay vs. Robert, 21.8%, 1625 cM, 23 seg (zero X-chromosome)
    Shay vs. Janice, 28%, 2085 cM, 24 seg (full X-chromosome, 182cM)
    Shay vs. Rich, 19.7%, 1466 cM, 24 seg (full X-chromosome, 182cM)
    Shay vs. Marcy, 30.1%, 2236 cM, 29 seg (zero X-chromosome)

    1/2 sister vs. paternal 1/2 brother, 26.6%, 1978 cM, 32 seg (share no X)
    1/2 sister vs. paternal 1/2 sister, 29.2%, 2176 cM, 35 seg (share full X, 182cM, 2.45%)
    1/2 sister vs. paternal 1/2 sister, 23.0%, 1713 cM, 31 seg (share full X, 182cM, 2.45%)

    Paternal aunt vs. paternal nephew, 26.7%, 1984 cM, 46 seg
    Paternal aunt vs. paternal nephew, 26.2%, 1946 cM, 46 seg
    Paternal aunt vs. paternal nephew, 24.9%, 1850 cM, 42 seg
    Paternal uncle vs. paternal nephew, 21.7%, 1611 cM, 46 seg
    Paternal uncle vs. paternal niece, 25.0%, 1857 cM, 55 seg (includes 2X seg, 81.6cM)
    Paternal uncle vs. paternal niece, 24.5%, 1823 cM, 53 seg (includes 2X seg, 81.6cM)
    Maternal nephew vs me 20.1%, 1495.4 cM, 35 seg (includes 1X seg, 18.5cM)
    Maternal nephew vs my full sister 28.6%, 2127.8 cM, 40 seg (includes some completely identical X seg, 169cM)

    • I miss stated (completely identical X-segment), I was thinking about his mother and my sister sharing a large amount of completely identical X chromosome.

  3. Hi I saw your question posed on Facebook but wasn’t sure how to complete form properly so I’m happy to give you GedMatch numbers so you can take a look if you like

    Grandson (M4+1) A347430 and his Maternal Grandmother (M0) A903999
    Note Grandson’s paternal great grandmother is Ashkenazi/endogamic. Grandson is 8% European Jewish according to Ancestry.

    Nephew (M4+1) A347430 and his Aunt (M2) A219400
    Nephew’s paternal great grandmother is Aunt’s paternal grandmother. She was Jewish.

  4. Sorry I’m of no help, but I share 811 with a great uncle and 942 with a great aunt. When my mother gets her results in a few weeks, we will know how much is shared between her and her aunt.

    I did have a quick question. On 23andme, my great uncle got a little (0.1%) Sub-Saharan African. I was wondering if all the ancestors he has, I also have, but he doesn’t have all the ancestors that I have? And if he has African ancestry from a slave (as he has deep ancestry is from Kentucky and Virginia), would I also have that ancestor, just no Sub-Saharan African showed up for me on ancestry?

  5. The maternal or paternal question has now been added to the bottom of the form
    Please any of you who still have the window open so can edit their response, add that information. Anyone who cannot edi,t send me the answer with the total cM and aprox time and date when you added your numbers, thanks!

  6. Done – this was an interesting exercise. I provided information for me and my two brothers compared with my mother’s sister and my mother with my mother’s sister’s daughter.

    I also added my half brother, but that is from Ancestry.com (he hasn’t gotten around to doing more with his DNA – hopefully soon I’ll get him on GEDmatch).

  7. This is very timely as I am trying to figure out if my FIL has found an aunt or a half-sibling to his match. They match at 2009 cM with 45 matching segments, the largest being 199.3 cM’s. FIL was b. 1947 and the match was b. 1938, so this is probably not grandparent/grandchild relationship.

    We did not know of this family, I was looking for a different paternal family.

    The match has a nephew (a son of her sister) who matches at 1548 cM and 42 matching segments with the largest being 100 cM. The nephew was b. 1973. It is assumed that the nephew and the aunt are full matches, not half, but the nephews mom has not yet tested.

    The nephew matches my FIL with 747 cM’s and 26 matching segments, the largest being 77 cM’s.

    I am going to fill out the Google document for each. Thank you.

  8. Dear Kitty,
    Thank you for this helpful blog post. I recently uploaded my DNA to Gedmatch from Ancestry. I am adopted and know my birthmother. My half-sister will have her DNA results soon. However,I had an interesting match come up and using your Google form, it appears that the gentleman in question is either a half-sibling, Uncle, or Grandparent. I was surpised given that our shared cM’s are not that high- 59.8. Perhaps I filled out the form incorrectly. Thank you in advance for any thoughts on this matter.

  9. Hi Kitty. I have entered two grandparent/grandchild and six aunt/uncle to nephew/neice for you! Will be interested to see the outcome of this data analysis.

Leave a Reply

Your email address will not be published. Required fields are marked *