Sunday, May 14, 2017

Partially Endogamous DNA

I've written a lot about my 100% Ashkenazi Jewish (AJ) heritage and how that means I share way more DNA with known cousins than would be expected.  But what about people who have some AJ heritage but also have ancestors with other backgrounds?  Do they still need to account for endogamy?  To see how partial endogamy impacts amounts of shared DNA with relatives on the endogamous side, I looked at relatives who have tested who are partially AJ (including those that are 3/4, 1/2, 1/4 and 1/8 AJ) to see how endogamy impacts DNA they share with others.

Analysis of shared DNA for those with partial Ashkenazic Jewish ancestry (in blue) with known relatives (details below)

Mark (3/4 AJ)

My third cousin Mark is 3/4 Jewish; all of his grandparents were Jewish with the exception of his maternal grandfather.  The following demonstrates how much DNA he shares with known relatives, as opposed to how much he'd be expected to share.
Name Actual Relationship Shared cM ISOGG average % above avg Comments
Eddie 2C1R 184.8 106.25 73.9%
Linda 2C1R 189.3 106.25 78.1%
Karen 2C1R 101.0 106.25 -4.9%
Leslie 2C1R 140.7 106.25 32.4%
Sandy 2C1R 165.4 106.25 55% 3/4 AJ
Ken 2C2R 180.7 53.13 240%
Marilyn 2C2R 162.0 53.13 201.1%
Myron 2C2R
53.13 76.4%
Sue 1/2 2C2R
26.56 -8.4%
Me 3C 108.8 53.13 104.8%
Sara 3C1R 75.9 26.56 185.8%

Unlike my own 100% AJ results, Mark's results are a bit more mixed.  He does have several matches that share significantly more DNA with him than would be expected--up to 240% in the case of his second cousin, twice removed Ken.  But there are a few cousins who share just a bit less than would be expected.

Sandy (3/4 AJ)

Sandy is my second cousin once removed.  Her grandparents are Jewish with the exception of her paternal grandfather.
Name Actual Relationship Shared cM ISOGG average % above avg Comments
Eddie 2C 169.1 212.5 -20.4%
Linda 2C 324.2 212.5 52.6%
Karen 2C 174.9 212.5 -21.5%
Leslie 2C 244.1 212.5 14.8%
Mark 2C1R 165.4 106.25 55.7% 3/4 AJ
Ken 2C1R 128.4 106.25 20.8%
Marilyn 2C1R 171.6 106.25 61.5%
Myron 2C1R
106.25 18.5%
Me 2C1R 131.9 106.25 24.1%
Sue 1/2 2C1R
53.13 68.5%
Sara 3C 71.8 53.13 35.1%

Sandy's results are even more mixed than Mark's.  Interestingly, while Sandy and my mother Linda share 52.6% more DNA than would be expected with their relationship, Sandy shares 20.4% less DNA with my uncle Eddie (Linda's brother) than expected.

Beth, Liz, Dave & Shannon (in blue) are partially Ashkenazi.  Below, their shared DNA with the yellow individuals is analyzed.

Beth (1/2 AJ)

Beth is my first cousin once removed.  Her mother was Jewish; her father was not, so her DNA is 50% Jewish.
Name Actual Relationship Shared cM ISOGG average % above avg Comments
Marvin 2C 334.7 212.5 57.5%
Sid 2C 337.0 212.5 58.5%
Susie 2C 363.5 212.5 71.1%
Berly 2C 183.3 212.5 -13.7%
Paula 2C 218.2 212.5 2.6%
Me 2C1R 214.0 106.25 101.4%

Beth shares more DNA than expected with my father (Marvin) and his siblings Sid and Susie.  But she shares close to the expected value with their first cousins Berly and Paula.  Perhaps this is because my grandmother shares quite a bit of DNA (in multiple small segments) with Beth--but maybe Paula & Berly's father didn't.  And I inherited much of the DNA which my father shared with Beth.

Dave (1/4 AJ)

Beth's nephew Dave had one Jewish grandparent paternal grandmother, so he has 1/4 Jewish DNA.
Name Actual Relationship Shared cM ISOGG average % above avg Comments
Marvin 2C1R 397.7 106.25 274.3%
Sid 2C1R 263.4 106.25 148.0%
Susie 2C1R 409.5 106.25 285.4%
Berly 2C1R 73.6 106.25 -30.7%
Paula 2C1R 63.1 106.25 -40.6%
Me 3C 300.7 53.13 466.0%

Here too, there are significantly larger amounts of shared DNA between Dave and my father and his siblings than between Dave and my father's first cousins.  And I share significantly more DNA with Dave than would be expected with a third cousin.

Liz (1/4 AJ)

Liz is a niece of Beth and a first cousin of Dave.  Her only Jewish grandparent was her maternal grandmother, so 25% of her DNA is endogamous.
Name Actual Relationship Shared cM ISOGG average % above avg Comments
Marvin 2C1R 167.6 106.25 57.7%
Sid 2C1R 145.7 106.25 37.1%
Susie 2C1R 265.2 106.25 149.6%
Berly 2C1R 134.8 106.25 26.9%
Paula 2C1R 169.4 106.25 59.4%
Me 3C 103.6 53.13 95.0%

Even though Liz has the same relationship to my branch of the family as her first cousin Dave, the amounts of DNA she shares with my father, his siblings, and his first cousins, are more stable than Dave (although Susie is somewhat of an outlier).  However, note that all of these matches are above what would be expected given the known relationships.

Shannon (1/8 AJ)

Liz's daughter Shannon has only one Jewish great grandparent (up her direct maternal line), so only about 1/8 of her DNA is endogamous.
Name Actual Relationship Shared cM ISOGG average % above avg Comments
Marvin 2C2R 117.1 53.13 120.4%
Sid 2C2R 89.6 53.13 68.6%
Susie 2C2R 143.0 53.13 169.1%
Berly 2C2R 69 53.13 29.9%
Paula 2C2R 76.9 53.13 44.7%
Me 3C1R 72.3 26.56 172.2%

Even though Shannon is only 1/8 AJ, endogamy still seems to be impacting how much DNA she shares with known relatives.  All 6 of her known relatives who have tested share more DNA with her than expected--sometimes quite a bit more.

So what conclusions can be drawn here?  Unfortunately things aren't as clear-cut as when one is dealing with 100% AJ DNA.  In some instances, the amount of shared DNA between an individual with partial AJ DNA and a known relative with full AJ DNA is actually quite close to the expected value.  But there are multiple examples--even when the AJ ancestor was several generations back--where a partially endogamous individual shares significantly more DNA with a relative than would be expected.

So if you have partial AJ ancestry, realize that large amounts of DNA may actually be due to endogamy--but not always.  Keep an open mind.

Note:  I'm on Twitter.  Follow me (@larasgenealogy).

Want to get future blog posts emailed to you automatically?
Enter your email address:


  1. Replies
    1. Thanks! You can probably do something similar (I'm not sure how many of your people aren't 100% AJ though).

    2. I hope to be able to use this principle to demonstrate something else entirely. When I get the results.

    3. Israel - looking forward to that. Ive been wanting to know the answer to this one for awhile and without enough data its very difficult to analyze.

    4. Interesting and very relevant to me (1/4 Jewish DNA), thanks. Several people on this page (i.e. many accounts managed by both you and IsraelP) match with me at a level that looks promising (e.g. 20-60cM), but who are probably highly distant relatives. But then again, my match with a known Jewish 3C1R is 9.7cM...

  2. yes thanks. I've been looking for something like this for awhile. Israel, looking forward to your reply as the more data the better. :)

  3. Fascinating, but what does ISOGG stand for? You inspire me to start making some charts . . .

    1. International Society of Genetic Genealogists

  4. I have just tested my two sons, who should appear as 50% Jewish. It will be interesting to see how they match up with the few known relatives of my husband who have tested.

  5. Lara, thanks for this. When you say "Even though Shannon is only 1/8 AJ, endogamy still seems to be impacting how much DNA she shares with known relatives," are you referring only to known AJ relatives, or to all relatives, AJ or not? Thanks again.

    1. Known AJ relatives--and I don't know her non-AJ lines, but I'd assume they'd not be impacted by endogamy.

    2. Thanks! Sorry I am seeing this so late. I appreciate your blog.

  6. Hi there! My mothers DNA came back with around 25% Ashkenazi Jew from Ancestry dot com. We have an unknown ancestor - her grandfather. We have a pretty good paper and DNA trail for her known family and there is no AJ there at all. So it seems highly likely the unknown ancestor is AJ. Ancestry dot also gave us two unknow second to third cousins. I've traced their ancestry back to a great grandmother. My question is though, can I trust that they are second cousins to my mother? They are second cousins to each other. The DNA is 300cms across 17 segments for one and 274 across 10 for the other. Thanks!

    1. Those are very good matches, regardless of endogamy. Have them transfer to GedMatch for free, and I bet you'll have several long segments. Second cousin seems about right!