Wednesday, December 11, 2019

The Vagaries of Autosomal DNA

I received Family Finder matches today for a known cousin. He is not yet part of my project so I cannot look at his matches with all the cousins together, but I did have a look at five individuals.

The first two are known first cousins to the match, and to each other.  The total cMs fall within the norm for first cousins according to Blaine's Shared cM Project.

The third match is with a half-nephew, which according to the Project is clustered with first cousins. But this match is significantly larger than the two first cousins, even a bit higher than the 95% percentile (1159 cM).

The fourth is a known second cousin - reasnable but slightly over the 95% percentile (397 cM).

The fifth is a bit of a surprise, as she is the sister of the second match at the top. Her match is more than 30% larger than his and this too is outside the 95th percentile. Her longest segment (94 cM) is about the s ame as her brother's (92 cM), so clearly there are significantly more segments. That longest segment is the same for both matches and is the segment we famously share with the descendants of Uncle Selig.

Each has segments that the other does not but clearly the sister has some very large ones, most notably 83 cM on chromosome 4, 49 cM on chromosome 11,  and 70 cM on chromosome 16.

Nothing here is out of the ordinary, but it gives me another opportunity to bang on the drum for testing everyone you can. Yes, siblings too.  They can be very different.


  1. Interesting. How do your cMs compare with Lara Diamond's ranges for Ashkenazis?

    1. Last time I looked, my own families fit her data well, but keep in mind that about eight hundred of her data points are from me.
      One of these days, I should give her some more.