For this first guest post, I’m pleased to welcome Lara Diamond, whose popular blog Lara’s Jewnealogy has been running since 2013. Endogamy has a very visible effect on chromosome mapping, as Lara explains very clearly below — Jonny
DNA Painter is a great tool for visualizing which parts of your DNA you inherited from which ancestors. As an example, if you and a known third cousin share specific segments of DNA, then you should be able to attribute those segments to your mutual great-great grandparents. If you have many DNA matches whose relationship is known to you, you can “paint” all the segments of your DNA that you have in common with any of those individuals to represent the known ancestors shared with each match.
And while in itself if pretty cool, DNA Painter can also help with placing new matches on the right branch of your family; if you’ve been able to attribute a segment of one of your chromosomes to a particular ancestor and a new match shares that segment as well, then they should be related on that branch.
However, when you’re from an endogamous population, it isn’t as straightforward (like pretty much everything dealing with endogamous genetic genealogy). In such a population, individuals are all descended from a small group of people whose descendants married within the group (and had children within that group) for generations—meaning that everyone is related to one another many times over. This is different than pedigree collapse; endogamy isn’t just known first and second cousins marrying, but it’s a situation where everyone is a cousin in many ways, often multiple generations back.
Examples of endogamous populations are Jews, Cajuns, Pacific Islanders and more—those limited to having children with partners from within their ethnic group because of religious, social, or geographic reasons. Basically, everyone in such a community is related to one another in multiple ways, so there is no choice but to marry someone who is a cousin to one degree or another.
People within these populations are therefore often predicted to be much more closely related than the paper trail demonstrates because the small bits of DNA that were inherited from many distant common ancestors can add up to a decent amount of shared DNA. Even if you know the relationship to a specific cousin, you’re likely related more distantly in multiple ways as well, which complicates matters. So when trying to attribute DNA to specific ancestors, most of the shared DNA will have come from known close common ancestors—but some will likely have come from other lines due to endogamous sharing.
Given all of this, what is an endogamous person to do? Does this mean that DNA Painter is unusable for endogamous populations? It definitely does not—but it means that there are some accommodations you will need to make in order to minimize the chance of falsely attributing segments to the wrong ancestors.
Endogamous users of DNA Painter will want to discount “smaller” segments when attributing shared segments to specific ancestors. While it is possible that a small segment is from the known shared ancestors, it’s also very possible that it’s an artifact of endogamy and was not a segment inherited from those known ancestors—so including them may well be incorrect and misleading.
DNA Painter’s default segment size to add to a map is 7cM. For those users not from an endogamous population, this is a reasonable minimum segment size, but that isn’t the case for endogamous users. For Ashkenazic Jews, I would recommend a minimum of 10cM segments—and even that will likely be mapping segments incorrectly to specific common ancestors. 15cM would minimize those incorrect segments—but the larger your minimum threshold, the more likely it is that you’ll be eliminating segments truly from common ancestors.
For a visual explanation of why you may want to discount those smaller segments (because visualization is what DNA Painter is all about!), we can take a look at my chromosome map. I’ve tested many relatives and identified others who tested independently, so I’ve been able to attribute quite a bit of my genome to specific ancestors on all lines of my family.
An example: “D,” an unknown match
The person highest on my GedMatch match list for which I don’t have a known connection is someone I’ll call D. D and I share 7 segments above 7cM, with 66.1cM shared in total. I can use DNA Painter to see which ancestors we may have in common based on overlaps with known matches whose shared DNA I’ve already added to my map. Since we all have two sets of chromosomes, DNA Painter can’t know if the match is on your maternal or paternal side—but D should be on one or the other.
In the typical (non-endogamous match), I’d expect to see DNA Painter narrowing down either the paternal or maternal attributions to one line of my family. But my paternal attribution includes both Diamonds (on my paternal grandfather’s line) and Garbers (on my maternal grandmother’s line). And the maternal attribution also covers ancestors of both of my maternal grandparents (and all four of my maternal great grandparents). These small shared segments with D (the largest of which is 11cM) are likely from multiple very distant ancestors and an artifact of endogamy—and probably not traceable.
A known relative example
You can also see this clearly looking at known relatives. My mom’s first cousin Karen (my first cousin once removed) is related to my mother through my maternal grandfather. My maternal grandparents’ ancestors are from geographically distinct areas. My grandfather’s ancestors were from the Russian Empire (on all lines at least back to the 1700s) and my grandmother’s ancestors were from Hungary (also at least back to the 1700s). So when I look at DNA that Karen shares with me, I’d expect to see the maternal attributions to all be ancestors of my grandfather. Except that Karen is (as am I) 100% Ashkenazic Jewish, so endogamy rears its ugly head.
We know that Karen is related to me on my maternal side, so I’ll concentrate on shared maternal segments. The Tolchins, Lefands, Supkoffs, Halperns and Lefand-Marienhoffs listed are all Karen’s documented ancestors (as well as those of my maternal grandfather). But the Rutners and Joshowitz couples listed are my grandmother’s ancestors—so not at all connected to Karen on any paper trail.
Now all those segments that Karen shares with me that are attributed to coming from my Rutner and Joshowitz ancestors are small—all in the 7-10cM range. (There are also some small segments in this range that are correctly attributed to ancestors of my maternal grandfather’s side.) If I didn’t know how Karen was related in advance, not painting these small segments for her would have made it immediately clear that she was related to my maternal grandfather.
So if you’re from an endogamous population, don’t discount DNA Painter. Its visualization capabilities are pretty cool, and concentrating on larger shared segments allows you to be more confident that you are painting segments for the correct ancestors. This can help you to understand how specific DNA matches could be related to you.