You may have heard the terms endogamy, pedigree collapse, or multiple relationships as you’ve been learning how to use DNA with genealogy. What does your match list look like if you have any of these scenarios? What do your cluster results look like? How do you work with those DNA matches? What does this mean for your research and analysis? This article will help you understand each term and give you examples to digest. I’ll also provide some high-level tips.
Let’s start with the least complicated of the three terms, Multiple Relationships. Put simply; this occurs when you are related to someone on more than one of your family lines, so you have multiple common ancestral couples. For example, my great-grandparents, Dock Harris and Alice Frazier, had siblings that married: Dock’s sister, Sophie Harris, married Alice’s brother, Millard Frazier. The diagram below shows three DNA matches who are my double third cousins. We have two sets of ancestors in common – John C. Harris & Malissa Welch (illustrated in pink) and Richard Frazier & Nancy Briscoe (illustrated in blue).
If a descendant of Dock and Alice Harris matches a descendant of Millard and Sophie Frazier, they will probably share more DNA than expected. This is because they have both the Harris/Welch and Frazier/Briscoe couples in common. An important step in analyzing DNA matches is to view the histogram for the proposed relationships using the Shared cM Project by Blaine Bettinger and hosted by DNA Painter. Let’s look at the histograms for the relationship of third cousin shown above. In the three images below, you can see the number of cM shared in red and an arrow pointing to where this amount fits with all of the submissions for third cousin. The numbers across the bottom of the histogram represent the amount of shared DNA, and the numbers running up the left side represent the number of people reporting that amount for the relationships of 3C. The numbers along the bottom represent the largest amount of shared DNA in that “bin” or range. For example, in the histogram below, the bin size is 25. In the bin for 100 cM, 689 people reported known third cousins sharing 76-100 cM.
How do we analyze these histograms? When viewing them on DNA Painter, we learn that the average amount of shared DNA (mean) for 3rd cousins in the Shared cM Project is 73 cM. All the values of 91, 105, and 108 are above the mean. With non-double cousins with independent descent lines, you would expect there to be more variation in amounts of shared DNA, with some sharing above and some below the mean.
Another measure to look at is whether the matches are within one standard deviation of the mean. The standard deviation for the 3C relationship is 43, so one standard deviation above the mean is 116 cM. All of these double 3C fall within one standard deviation, so the extra amount of DNA from the two shared ancestral couples wasn’t enough to make them an obvious outlier.
The conclusion is? Double relationships may not affect the DNA so much that it changes the relationship estimate, especially if it’s third cousins or more distant. However, if someone shares more than just two relationships, it may have a greater effect on the shared DNA.
One of my Frazier cousins shared her DNA with us. She is two generations closer to our Frazier shared ancestor. She has a DNA match on our Frazier side who appears at first glance to be a 3C1R through the Frazier/McChristian line. After closely analyzing their trees, Nicole found they actually share 4 common ancestral couples: John Clanton & Nancy, John Briscoe & Margaret Looney, Robert McChristian & his wife, and James Frazier & Nancy Lane. This makes him a quadruple cousin – a 3C1R three times and a 4C one time. He and my Frazier cousin share 153 cM, which is on the very edge of possibility for a 3C1R, as shown below. If you saw this in your own match list, you would want to explore the DNA match’s pedigree further and see if there are multiple relationships you hadn’t noted.
Our goal of genetic genealogy is to discover the common ancestor between a test-taker and a DNA match. We often accomplish this by clustering our matches using the DNA company shared matching tools or third-party tools.
How can multiple relationships affect your analysis of clusters? You may be confused looking at shared matches because they will match on two or more different lines. As you’re trying to separate your matches into clusters descending from a common ancestral couple, you may not realize that you have more than one connection to a match. When you see an anomaly such as this, step back and do more genealogy. In my case, I knew that siblings married siblings, and I could draw out the connections quite easily.
Let’s look at a network graph that Nicole created for me with Gephi. Network graphs show DNA matches as nodes or points and shared match connections as lines between two nodes. Groups of matches that have many connections to each other are called clusters or “communities” in Gephi. These clusters usually mean that the group of matches share an ancestral line – many of them the same MRCA couple (Most Recent Common Ancestral couple). (Review our other posts about network graphs here.)
To understand the network graph, look at my pedigree chart below. This is the FAN chart from FamilySearch showing my grandparents and great-grandparents. My four great-grandparent couples are Shults/Royston, Harris/Frazier, Kelsey/Beddoes, and Creer/Peterson.
The network graph below was created using a subset of my DNA matches at AncestryDNA, sharing from 35-268 cM. My two maternal great-grandparent couples are represented by clusters 0 (Kelsey/Beddoes) and 1 and 2 (Creer/Peterson). Descendants of my paternal great-grandparents are represented in clusters 3 (Shults/Royston and 4 (Harris/Frazier). Cluster 5 represents a common ancestral couple further back on the Royston side, Royston/Weatherford (parents of Dora Algie Royston). Because cluster 5 represents a common ancestral couple further back on the Shults/Royston side, you can see that there is a connection between cluster 5 and cluster 3 with a larger gold node and many connecting lines. In this network graph, matches sharing more DNA are represented by larger nodes. If more matches were included in this graph, you would likely see more connections between that cluster.
My Harris/Frazier double cousins are included in the green cluster, #4. An interesting match is the node labeled “M” in the center of the graph. My maternal ancestry is comprised of mid-1800 emigrants from England and Denmark. My paternal ancestry represents my Dad’s southern colonial roots. “M” is the only DNA match I’ve discovered who connects to both my paternal and maternal lines. We are double cousins and share one ancestral couple on my maternal side, Creer/Morris, and another ancestral couple on my paternal side, Royston/Weatherford. The result of this double cousin in the network graph is that clusters 1 and 5 appear to be connected, which usually indicates the MRCAs of the clusters are related. But in this case, it just shows that “M” matches both clusters.
Takeaways and Tips for Multiple Relationships
- Be careful not to confuse multiple relationships with pedigree collapse.
- Build complete trees for DNA matches to check for multiple shared ancestors.
- Evaluate each DNA match with the Shared cM Project Tool and explore inflated amounts of DNA.
- The amount of shared DNA will be affected more in close relationships than in distant ones and in cousins who share many MRCAs.
Next, let’s examine pedigree collapse and see what it looks like in DNA results. Pedigree collapse is different than multiple relationships. In my example above of double cousins, I didn’t have pedigree collapse; I just shared two sets of common ancestors with some of my DNA matches. Pedigree collapse occurs when cousins reproduce, and their descendants have ancestors in more than one place in the family tree. Pedigree collapse can become endogamy if it repeatedly occurs over many generations. However, two to three instances do not qualify the situation as endogamy.
How does pedigree collapse affect DNA matching? It can affect the amount of shared DNA with closer matches, but the effects fade quickly because of segment loss in each generation. The pedigree collapse will not have a significant effect on downstream matching. For example, the diagram below shows John Wesley Clark and Mary Ellen Locklear as both Donna’s paternal second great-grandparents and her maternal great-grandparents. Donna’s parents, Mary and John, are first cousins on the maternal side and first cousins once removed on the paternal side.
What does a DNA match look like with pedigree collapse? In the diagram below, you can see that Joan is a second cousin once removed on the paternal line and a second cousin twice removed on the maternal line.
You might expect Joan and Donna to have an inflated amount of shared DNA, so you’d want to check the Shared cM Project Tool. Viewing the histograms for the two relationships shows that sharing 225 cM for either relationship is on the far right shoulder, meaning that the shared DNA is on the high end of what’s expected. The 2C2R relationship especially is far from the mean of 71 cM, and higher than one standard deviation above the mean (113 cM).
If you see a DNA match sharing on the high end of what’s expected for their relationship, consider that they might have another path back to the same common ancestral couple.
What does pedigree collapse look like in a network graph? In the image below for Donna’s DNA matches, you can see how the clusters are very connected and don’t separate into distinct groups.
With pedigree collapsing, teasing out how you are connected to another DNA match becomes more complicated. In Donna’s case, the impact is quite severe because her maternal great-grandparents are also her paternal second great-grandparents. With segments lost in every additional generation, the pedigree collapse would not affect Donna’s descendants as much.
Another factor in Donna’s ancestry is her Native American heritage which had considerable intermarriage for generations. Donna’s DNA not only has the recent pedigree collapse but could also be impacted by some endogamy on those lines.
Takeaways and Tips for Pedigree Collapse
- The effect will be noticeable for close descendants of the collapsed lines.
- DNA segments are lost with each generation, so downstream matching won’t be largely affected.
- If you have pedigree collapse in your family tree, carefully evaluate each DNA match that comes through those lines.
- Recognize that it will be more difficult to cluster your DNA matches, but with careful diagramming and good genealogy, you can tease out the relationships.
Finally, let’s examine endogamy. Pedigree collapse over many generations can result in endogamy, but two to three instances do not. People reproducing within the same populations for hundreds of years – generations and generations – results in endogamy. This typically manifests in groups intermarrying for cultural, religious, or geographic reasons. Endogamous populations include Ashkenazi Jewish, Pacific Islanders, French Acadians, and Amish.
What is the effect on genetic genealogy? DNA matches will share many small segments making their total DNA larger, so they appear to be a closer cousin when they are actually a distant cousin with many shared common ancestors far back in time. Attempting to predict relationships becomes challenging with so many of the segments inherited from the endogamous population.
In the network graph below of an individual with Native American and Hispanic ancestry out of New Mexico, you can see two endogamous groups that share DNA in many ways. This individual has over 100,000 DNA matches on Ancestry, with 72,000 of them distant (sharing 6-20 cM). There were so many shared match connections that it was difficult to download the matches and ICW (in common with) using the DNAGedcom Client, so a smaller range of just 68-90cM was used to generate this network graph. As you can see with all the connecting lines, it appears that everyone shares with almost everyone else.
As you can imagine, attempting to discover a common ancestor without distinct clusters is very difficult in an endogamous population. Some considerations – first, you’ll want to look at your closer matches, say over 200cM. Also, look at the size of the segments. You’ll want to focus on matches where the largest segment is at least larger than 20 cM and maybe even 30 cM. Larger segments mean that you are likely related more recently than very far back. Also, learn how each testing company deals with endogamy. Each company has its own strategy, and its white papers can explain the science behind its algorithms.
Roberta Estes’ blog post “DNA: In search of. . . Signs of Endogamy” discusses endogamy in detail and details many strategies for using ethnicity and segment data. She also includes a summary of endogamy tools at each DNA testing website. Many others have written about endogamy, and I’ve included a list of resources for you to peruse at the end of this article.
Takeaways and Tips for Endogamy
- Pedigree collapse can turn into endogamy, and it can be difficult to know where one ends and the other begins.
- With endogamy, you can’t cluster matches, and you might be dealing with inflated amounts of shared DNA that don’t reflect the correct relationship. This is not always the case.
- You could have many, many matches, so work with the best matches – those sharing more DNA with larger segments.
- Testing other relatives is also beneficial to finding good matches.
- Consider Y-DNA and mitochondrial testing for relatives, which can help sort out various family lines.
- Finally, you’ll need to build trees for the best matches.
Diahan Southard offers an online course about dealing with endogamy. Diahan has a wonderful way of teaching about DNA that helps you gain confidence! Learn more and register here: Start Untangling Your Family Tree | Endogamy & DNA Course
To learn how to create network graphs like the ones in this article, join our online course, Research Like a Pro with DNA. In this course, we take you step by step through organizing and analyzing your matches and applying DNA evidence to focused research questions.
Diamond, Lara. “Chromosome Mapping and Endogamy.” DNA Painter Blog. https://blog.dnapainter.com/blog/chromosome-mapping-and-endogamy/. Posted 30 January 2022.
Ekins, Jayne. “Pedigree Collapse and Your DNA Matches.” Your DNA Guide. https://www.yourdnaguide.com/ydgblog/pedigree-collapse-and-genetic-relationships.
Estes, Robert. “Concepts – The Faces of Endogamy.” DNA eXplained – Genetic Genealogy. https://dna-explained.com/2017/03/10/concepts-the-faces-of-endogamy/. Posted 10 March 2017.
Estes, Roberta. “DNA: In Search of. . . Signs of Endogamy.” DNA eXplained – Genetic Genealogy. https://dna-explained.com/2022/08/11/dna-in-search-ofsigns-of-endogamy/. Posted 11 August 2022.
Estes, Robert. “What’s the Difference Between Pedigree Collapse and Endogamy?” DNA eXplained – Genetic Genealogy. https://dna-explained.com/2021/07/23/whats-the-difference-between-pedigree-collapse-and-endogamy/. Posted 23 July 2021.
International Society of Genetic Genealogy Wiki. “Endogamy. Latest update 10 July 2021.
Larkin, Leah. “The Endogamy Files: What is Endogamy?” The DNA Geek. https://thednageek.com/the-endogamy-files-what-is-endogamy/. Posted 5 May 2020.
Powell, Kimberly. “The Challenge of Endogamy and Pedigree Collapse.” Advanced Genetic Genealogy: Techniques and Case Studies. Editor, Debbie Parker Wayne. Iron Gate Publishing, 2019.
Woodbury, Paul. Dealing with Endogamy, Part 1: Exploring Amounts of Shared DNA.” LegacyTree Genealogists. https://www.legacytree.com/blog/dealing-endogamy-part-exploring-amounts-shared-dna. Posted November 2016.
Woodbury, Paul. Dealing with Endogamy, Part 2: Test Multiple Relatives.” LegacyTree Genealogists. https://www.legacytree.com/blog/dealing-endogamy-part-ii-test-multiple-relatives. Posted November 2016.