Sunday, April 10, 2016

How Endogamy Looks in Practice

(Note that there is a more updated version of this post here with the benefit of another year's worth of known relatives testing.)

In response to recent posts about how I'm related to people in multiple ways, and why this expression of Ashkenazi Jewish endogamy makes using DNA to find relationships among those with Jewish ancestry much more difficult, an adoptee asked how far off from how much shared DNA would be expected my known relatives were in order to help her understand her own matches better.

We always hear about how much harder it is to use genetic genealogy to find relatives when you are from an endogamous background, and Ashkenazi Jews (AJs) are a typical example.  AJs married within the small pool of AJs for centuries, so we are all related to one another in multiple ways.  I've had enough relatives test to see how endogamy is reflected in practice.
Some of my known relatives who have tested, displayed in FamilyTreeDNA

I knew that I'd probably share more DNA with my known relatives than would be expected, but the amounts that we actually shared, especially with somewhat more distant cousins, shocked me.

Name Actual Relationship FTDNA's Predicted Relationship Range Shared cM ISOGG average % above avg Comments
Paternal Grandmother Grandmother Half Siblings, Grandparent/ Grandchild, Aunt/ Uncle, Niece/ Nephew 1327.1794 1700 -22%
Paternal uncle Uncle Full Siblings, Half Siblings, Grandparent/ Grandchild, Aunt/ Uncle, Niece/ Nephew 1955.2031 1700 15%
Maternal uncle Uncle Half Siblings, Grandparent/ Grandchild, Aunt/ Uncle , Niece/ Nephew 1848.0041 1700 8.7%
Ruth Great Aunt 1st Cousin, Half Siblings, Grandparent/ Grandchild, Aunt/ Uncle, Niece/ Nephew 1008.4038 850 18.6%
Scott 1C1R 1st Cousin, Aunt/ Uncle, Niece/ Nephew 685.0191 425 61%
George 1C2R 1st Cousin - 3rd Cousin
212.5 27.4%
(George's 1/2 1C)
Half 1C2R 2nd Cousin - 3rd Cousin 206.0462 106.25 193%
Ken 2C2R 2nd Cousin - 3rd Cousin 108.1761 53.13 103%
(2nd cousin to Ken; 1/2 second cousin to Sue)
2C2R 3rd Cousin - 5th Cousin 174.8590 53.13 229%
(Myron's niece)
3C1R 3rd Cousin - 5th Cousin 89.1355 26.56 235%
(1/2 second cousin to Ken and Myron)
Half 2C2R 2nd Cousin - 4th Cousin 103.5133 26.56 289%
Beth 2C1R 2nd Cousin - 3rd Cousin 232.1945 106.25 118% 2 AJ grandparents
(Beth's nephew)
3C 1st Cousin - 3rd Cousin 276.0552 53.13 419% 1 AJ grandparent
Judith 3C1R 2nd Cousin - 3rd Cousin 186.3940 26.56 602%
Jonathan (Judith's son) 4C 2nd Cousin - 3rd Cousin 182.0750 13.28 1271%
Pat 2C1R 2nd Cousin - 4th Cousin 117.5067 106.25 10.6%
Ben 3C, 4C 3rd Cousin - 5th Cousin 101.6712 53.13+ ~91% Has 2 Supkoff great grandparents
Table Showing Actual & Genetic Relationships Between Myself and Known Relatives

In the above chart, the columns represent the following:
  1. The first column is the relative who has tested
  2. The second column is the testee's known relationship to me; xCyR means that the person is an xth cousin, y times removed.
  3. The third column is FamilyTreeDNA's predicted relationship range between myself and the relative.  Note that FTDNA does try to adjust this somewhat for those with Jewish DNA.
  4. The fourth columns is how many cM of DNA I share with the individual
  5. The fifth column is how much I'd be expected to share with that individual, per ISOGG's Autosomal DNA statistics.
  6. The sixth column is the percentage of cM shared above what would be expected for a specific relationship.
  7. Finally, the seventh column has additional information about the individual which may be relevant.
Unless specifically noted, all of the individuals on this list had 4 Ashkenazi Jewish (AJ) grandparents.

With the exception of my grandmother (I guess I got more DNA from my paternal grandfather than grandmother), every individual on this list shares significantly more cM with me than would be expected for their known level of cousinship.  Even Dave, who has only 1 Jewish grandparent, shares more than four times the expected amount of DNA with me.

The "winner" in the endogamy contest is my fourth cousin Jonathan, who shares 182cM with me, as opposed to the expected 13.28cM.  While his mother (Judith) already shares much more DNA with me than would be expected, Jonathan shares almost as much, even though he's another generation removed.  So while at first glance, it would seem he just inherited the majority of what his mother shares with me, I took a closer look.
Some of the segments I share with Judith & Jonathan
In some cases (such as on chromosome 18 and a small one on chromosome 19), segments shared by Judith and myself were passed onto Jonathan intact.  In others (like on chromosome 6), a relatively large segment which I share with Judith wasn't passed to Jonathan at all.  But then there are the segments which I share with Jonathan on segments 19 and 7--segments which I do not share with Judith, so they came from Jonathan's father (to whom I have no known relationship).  This is endogamy in action and shows that some of the segments that Ashkenazi Jews have in common with known relatives are quite likely from other ancestors than the known shared ones.

Hopefully this will give more insight into what different values for shared cMs can really mean when dealing with Ashkenazi ancestry.  And good luck.  You'll need it.

Note:  I'm on Twitter.  Feel free to follow me (@larasgenealogy).

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


  1. Thanks Lara....I too have some strange examples - two of my brother's children are 128 and 121% respectively but my sister's son is 89.9%. And my two first cousins (from maternal siblings) are 105% and 72%. It seems I match the males from both sides of my family more closely. Now to get more family to test.

  2. When a known relative is a third or fourth cousin or even further, it does not require a lot of smaller more distant relationships to kick then up a level.

    When a closer known relative has much more than expected, there is probably some fairly close, though unknown,relationship lurking in the background.

    It is important to distinguish between these two things - general endogamy and missing information. They are very different phenomena.

  3. This is a couple of years old but packed with information. Fig 4 and Fig 5 are particularly interesting - showing the extra shared DNA Ashkenazim have with cousins of the same distance as compared to the same relationships among non-Jewish subjects.

    For me, the biggest surprise was where all the extra DNA was coming from. I knew there were more segments but I expected the "extra" segments to be mostly under 3 cM. They were actually in the 3 to 5 cM and 5 to 8 cM range. See Tables 8 and 9.

  4. Thank you! I look forward to getting more of my (AJ) husband's relatives to test. I hope I can ultimately generate as interesting blog post as this one!

  5. I've compared your values to the ranges plotted for the shared cM project ( and and Paul Rakov's simulations ( This shows that most of these distances are within the predicted range of variation for the relationship. I think only Dave, Judith and Jonathan have unexpected amounts given the relationship. The more interesting point is that despite this, all but one relationship here shows more than the average sharing. The additional amount above the average is roughly 50-250 cM apart from Sue. This suggests to me that this is the level of background sharing to be expected amongst AJ, and one should be allowing for this in predicting cousin levels. The paper Gaye linked to in surnamedna journal shows similar levels of excess sharing. I think it is the absolute cM extra that matters not the percentage, and that the paper misses a trick by not calculating that.

  6. my QUESITION. ... i "found" in Prague "my relative cousine" ... from family Schalek line separated since 1787 when my ancestor Abraham Schalek changed name to Travnicek .... how relevant could be DNA testing when we could try it ? .. the gap between me and his tree is maybe since 1720-30 if Schaleks were brothers or relative
    (no chance to do other way because Familiant record has no info about his Schalek's Familiant parents)

    any comments are welcomed :) ... better to send to my email

    1. Assuming you're Jewish, it'll be difficult to know if any shared DNA is because of your Schalek ancestor or from another line that you haven't traced that far--and with Jewish endogamy, odds are that they're related another way as well.

  7. Lara,

    I want to let you know that your blog post is listed in today's Fab Finds post at

    Have a great weekend!

  8. You and I share more centimorgans than do and Jonathan (FTDNA). Funny how that works!

  9. I posted that you and I share more centimorgans than Jonathan but what I meant was that we share more "projected" centimorgans. We share 45 (ft DNA)