We all have pedigree collapse in our trees. You may not have found it in yours but you will. Pedigree collapse is when two of your ancestors descend from the same ancestor or ancestral couple. Sometimes called “cousins marrying cousins” pedigree collapse is not the same as endogamy. Endogamy is when cousins are marrying cousins over multiple generations. I have found pedigree collapse on both my paternal & maternal sides. I have to confess. When I first discovered this in my lines I was excited. It meant that there where one or two less ancestors I had to research. Two for the price of one. It doesn’t get any better. Then DNA came along and now that same pedigree collapse I was so excited about, well let’s just say it can make figuring out how I am genetically related to a match a wee bit more complicated.
DNA is a piece of evidence and nothing more. It must be used with other pieces of evidence. In short DNA does does not stand on it’s own. Without paper trails you can not determine how you are related to a DNA match. All of the major DNA testing companies used for genealogy give a relationship range to a match. Those relationships are primarily based on how much DNA two kits share and believe me it can very widely. Once you are beyond parents and full siblings they are predictions, best guesses and nothing more. The amount of DNA shared between two matches is represented in centiMorgans (cMs). A cM is a unit for measuring genetic linkage. The more cMs you share with a match the closer your relationship. Below is a chart compiled by Blaine T Bettinger from data collected for The Shared cM Project which shows just how widely the amount of shared DNA varies for various relationships.1
The pedigree collapse on my paternal grandfathers lines originates in the U.S. with a group of immigrant ancestors who arrived in 1738 and 1739. In 1764 they were finally awarded the land they had been promised when they first left Islay, Scotland. Known as The Argyle Patent, the land is located in Washington County, New York.
The first DNA test I was involved with was a yDNA test.2 It was July 2007. I was standing on the west coast of Scotland gazing out across the North Atlantic Ocean contemplating research strategies. A few days earlier I had been to the National Archives of Scotland located in Edinburgh. Records from Islay and more specifically from the early 1700’s were not readily available. Of the records available they were spotty at best. Were we ever going to figure out how McIlepheder became McNeil? It occurred to me DNA might help. Standing on that cliff as the waves crashed against the shore I pulled out my mobile, as they are called in the U.K., and called a McNeil cousin in the States. “The only way we are going to stand a chance of figuring out the McNeil name change is by doing a yDNA test. Do you think your dad would do one?” “I think so. I’ll ask him.” That day began my journey into DNA as a piece of evidence. It took a number of years before a we got a match and we are still waiting for a match from Scotland to help answer the original question. In 2011 I did my first autosomal DNA (atDNA) test. The following year I tested my mom and dad. Now I manage 30+ DNA kits from various family members.
DNA and Pedigree Collapse
The First Match
Dad’s results came back. His and my McNeil cousin, Angus (alias for privacy) had an atDNA match. The match on chromosome 1 is from 202 to 223.5 for a total distance of 23.8 cMs.3 While the match does indicate my dad & Angus are related genetically and to a certain point the test adds an infinitesimal amount of weight to the paper trail. It does not tell me how they are genetically related. You see Dad & Angus share more than one common ancestor couple and dad has pedigree collapse on the related lines. See the circle chart below.
A word about the chart.
This chart is a modified version of a descendant circle chart. Traditional versions of this chart have a single ancestral couple in the center with the descendent’s radiating out. Each concentric circle represents one generation from the ancestral couple. I have used the same concept except the center contains four ancestral couples. The two in white are the ancestral couples Dad and Angus share. The other two are two separate Gilchrist lines which may or may not be related. All of the ancestors represented by the gray circles are known and have been throughly researched. I have chosen to leave them off the chart in order to focus on the relevant ancestors. No other overlap has been found between the two trees.
Dad’s paths of Inheritance.
1. [—? —] McIlepheder > daughter Catherine > great-grandson Alexander R Gilchrist > son John Duncan Gilchrest > grandson Dad.
2. [—? —] McIlepheder > son Archibald > 2X great-granddaughter Margaret Ella White > grandson > Dad.
3. James Gillies and Ann Campbell > daughter Margaret Gillies > granddaughter > Margaret Gilchrist > son John Duncan Gilchrest > grandson > Dad.
Angus’ paths of Inheritance
1. [—? —] McIlepheder > son Archibald > son Alexander McNeil > 3x great-grandson > Angus.
2. James Gillies and Ann Campbell > daughter Catherine Gillies > 3x great-grandson > Angus.
The closest relationship between Dad and Angus is double fifth cousins (5C). They also are 6C.
Angus’ MRCA (most recent common ancestor) couple is Alexander McNeil and Catherine Gillies, his 3x great-grandparents. In other words Angus received his segment from either Alexander McNeil or Catherine Gillies. Dad’s MRCA couple are his grandparents John Duncan Gilchrest and Margaret Ella White. Why, you ask? That pesky pedigree collapse. John and Margaret are 3C1R (third cousins once removed). None of the three paths Dad could have inherited the DNA can be eliminated from consideration. The only thing I know is he received it from one of his paternal grandparents. The goal is to determine which grandparent the segment came from. Luckily there is a technique known as visual phasing that can help determine which grandparent, John or Margaret passed the segment to Dad, potentially eliminating at least one of the paths of inheritance.
Each person inherits about 25% of their DNA from each one of their four grandparents. Visual phasing allows you to determine which pieces came from which grandparent. One of the biggest benefits of visual phasing is you can eliminate a quarter of your tree when searching for a common ancestor between two matches. The technique is best done with three siblings. It can be done with two siblings but requires a different approach and a lot of second or greater cousins. Dad had two brothers, one died well before atDNA testing was available. The only option is two sibling visual phasing. In 2015 Dad’s brother, Ewan’s (alias for privacy) test came back. The first step to visually phase a chromosome is to compare the kits against each other. In three sibling visual phasing you are doing three comparisons. With two siblings you only have one comparison. The second step is identifying the crossover/recombination points. Once the recombination points are identified and you have three siblings you can determine which sibling owns/has the recombination. The only way to determine which sibling owns/has the recombination point with two siblings is by the use of cousin matches. Below is the one to one comparison between Dad and Ewan on chromosome one showing the section which contains the match Dad and Angus share.
Legend Green = Fully Identical. Yellow = Half Identical. Blue = Share DNA
DNA comes in pairs. You inherit one side of the pair from your father and the other side of your pair from your mother. These pairs/halves are a recombination of the two halves each parent has. With this understanding a fully identical segment means that Dad and Ewan received the exact same piece of DNA from two of their grandparents, one on the paternal side and one from the maternal side. Half identical means that Dad and Ewan received the exact same piece from one grand parent either paternal or maternal. The other half of the pair of chromosomes came from opposite grandparents on either the paternal or maternal side. An example of fully identical would be Dad and Ewan inherited the same piece from there paternal grandfather and the same piece from their maternal grandmother. An example of half identical would be Dad and Ewan received one of their halves from their maternal grandmother. The other half they received from their father but Dad inherited his paternal grandfathers DNA and Ewan inherited his paternal grandmothers DNA.
End Side Bar
Dad’s and Ewan’s one to one comparison shows recombination points at 199 and at 232.6 and they are half identical between those two points. The segment (202.3—223.5) Dad and Angus match on lies between those recombination points. A comparison of Ewan to Angus reveals they also match from 202.3—223.5.4 Because both Dad and Ewan match Angus and I know Angus is a paternal match. I can say that Dad and Ewan inherited the same piece of DNA from one of their paternal grandparents. Because of pedigree collapse I don’t know which one. On the maternal side Dad either inherited his DNA from is maternal grandfather or grandmother. Ewan’s DNA came from which ever one Dad did not get his from. Remember Dad & Ewan are half-identical. Since I don’t have any other cousin matches on this segment I still don’t know which grandparent on the paternal side passed on the matching segment. Drat that pedigree collapse! You are not my friend.
The Second Match
In 2017 I uploaded a number of DNA kits to MyHeritage. This year a new match, Fiona (alias for privacy) appeared for Dad and his brother Ewan. Dad matches Fiona on three segments for a total of 38.5 cMs. While Ewan matches on two segments for a total of 25.6 cMs.
As soon as I downloaded the match data and imported it into Genome Mate Pro (GMP)5 I was doing the happy dance. Why, you ask? One of the segments Dad, Ewan and Fiona match on just happens to be on chromosome one and it is in the same place as Angus’ match to Dad and his brother. Once again I was on the cell phone calling Angus’ son, “I need you to transfer your Dad’s DNA to MyHeritage. Then see if he matches Fiona and if so download the segment data and send it to me.” “OK, I should be able to do that in the next day or two.” “What’s up?” “I think we may have triangulation. I need your Dad’s and Fiona’s match data to know for sure.” This was before MyHeritage released the chromosome browser that does one of the requirements of triangulation for you.
Matching segments from GMP. Match data for Angus is from Family Tree DNA (FTDNA). Fiona’s match data is from MyHeritage.
What is Triangulation?
Triangulation is a rigorous method of analysis used in genetic genealogy. There are three requirements that must be met in order to say a segment is triangulated. They are:
- There must be a minimum of three tests/legs.
- All three tests must match each other on the same segment and the segments must overlap each other.
- The legs cannot be closely related. For example all of the following relationships are considered one test/leg, parent/child, aunt/uncle, two siblings, grandparent/grandchild, and first cousin relationships.
Do We Have Triangulation?
Since Fiona is not on GEDMatch or FTDNA I have used the segment data from MyHeritage for all of the following comparisons.6 The only difference between MyHeritage and GEDMatch is MyHeritage provides slightly different start and stop points and the total shared cMs is slightly longer 25.7 verses 24.3 for Dad’s and Angus’ match. These differences are not significant. Start and stop points are fuzzy. The primary reason is due to how the various companies round the start and stop point.7 The final difference is MyHeritage uses imputation8 for the number of SNP’s (snips) they are comparing. GEDMatch uses only the SNP’s that are actually tested. The effect is the number of SNP’s from MyHeritage is more than double what they are on GEDMatch.
What does this actually mean?
Ideally it would have been nice if Dad’s and Ewan’s match to Fiona was larger than 10 cMs. Why? The biggest reason is the larger the segment the less likely it will be a false positive or IBS (Identical By State). An IBS segment means the segment is identical but the match is not identical by inheritance/descent (IBD) from a common ancestor. Phasing a test to parents decreases the likelihood that a match is IBS.9 While I can’t phase Dad and Ewan’s DNA to their parents, one of my siblings, Blair (alias for privacy) also matches Fiona and Angus on chromosome 1 on the same segments Dad matches Angus and Fiona. Blair’s segment visually phases to Dad’s father, Blair’s paternal grandfather. One of Blair’s children, Liam (alias for privacy) matches both Fiona & Angus on the same segments. A second child Eoghan (alias for privacy) matches Angus but not Fiona. A comparison of Liam to Eoghan reveals a recombination point at 216 explaining why Eoghan does not match Fiona. Blair’s, Liam’s and Eoghan’s phased DNA matches to Fiona increases the likelihood Dad and Ewan’s match to Fiona is IBD. Triangulation of the segment will dramatically increase the likelihood that the segment is IBD.
The criteria for triangulation requires 3 standards be met. So far the first two have been met. Three tests match and overlap on the same segment. Dad and Angus are not close relatives as can be seen in the descendant circle chart above. Is Fiona far enough away from both Dad and Angus to met the third standard?
Descendant Circle Chart Dad and Fiona.
Determining Dad’s and Fiona’s common ancestors is a bit more complicated than Dad and Angus. As you can see by the chart there are a plethora of ways Dad’s and Fiona’s matching segment could have been passed down. All of them need to be identified.
Dad’s possible paths of Inheritance
1. John McDougall > great-granddaughter Margaret White > grandson Dad.
2. John Gilchrist > Duncan J Gilchrist > grandson John Duncan Gilchrest > grandson Dad.
3. Margaret Gillies > to Duncan J Gilchrist > grandson John Duncan Gilchrest > grandson Dad.
4. Margaret McCaslin > grandson John Duncan Gilchrest > grandson Dad.
Fiona’s possible paths of Inheritance
1. Mary Ann McDougall > Catherine McCollum > James Henry Gilchrist > great granddaughter Fiona.
2. John Gilchrist > Duncan J Gilchrist > grandson James Henry Gilchrist > great granddaughter Fiona.
3. Margaret Gillies > Duncan J Gilchrist > grandson James Henry Gilchrist > great granddaughter Fiona.
4. Margaret McCaslin > grandson James Henry Gilchrist > great granddaughter Fiona.
The MRCA couple between Dad and Fiona are Duncan J Gilchrest and Margaret McCaslin making them third cousins once removed (3C1R). Dad and Fiona are far enough apart. Are Fiona & Angus?
Descendant Chart Angus and Fiona.
Angus and Fiona are fourth cousins twice removed (4C2R). The MRCA couple are James Gillies and Ann Campbell.
We Have Triangulation.
The three requirements for triangulation have been met.
1. Do we have three tests/legs? Yes. Dad, Angus and Fiona.
2. Do all three kits match each other and overlap. Yes.
3. Are the tests/legs far enough apart? Yes.
Dad and Angus are 5C & 6C.
Dad and Fiona are 3C1R.
Angus and Fiona are 4C2R.
All of the evidence is clearly indicating the match is IBD.
Which line did the DNA segment come from?
When comparing all three relationship charts there is one and only one couple common to all three trees. James Gillies and Ann Campbell. However not all three tests can assign the exact same segment to James Gillies and Ann Campbell. Segment assignment depends on which kit you are looking at.
The gold segment Dad and Ewan share from 199 to 232.6 can be assigned to their paternal grandmother Margaret Gilchrist. She received the segment from one of her parents. This segment is now visually phased. I still don’t know who owns the recombination points. Part of Dad’s and Ewan’s matching segment, the part from 204 to 226 (purple) can be assigned to James Gillies and Ann Campbell. This segment contains Dad’s and Ewan’s match to Fiona.
Angus can assign the purple segment plus the green segment to James Gillies and Ann Campbell. Because of the match with Fiona Angus is able to step the segment he shares with Dad back one more generation.
Fiona can assign the green segment to James Gillies and Ann Campbell.
Friend or Foe?
Depending on what I am working on pedigree collapse is still my friend but when I am using DNA as a piece of evidence it can very quickly become my foe. As can be seen here Dad had three paths he could have inherited the DNA from with his match to Angus and four from his match to Fiona. Without both Angus’ and Fiona’s match there was no way to determine how Dad inherited his segment.
A word of caution.
DNA evidence is no different than any other evidence in genealogy. When you present a case study, proof argument or any assertion in genealogy. New evidence has the potential to support or refute previous conclusions. All three trees have been well documented on the matching lines back to James Gillies and Ann Campbell. However that does not preclude the possibility of another in-common line showing up at some future date. Every new match that Dad has on this segment will need to be evaluated using the same techniques to determine if the conclusions hold up.
- The Shared cM Project is a project where people who have done DNA tests and have known matches can submit how much DNA a specific relationship shares. The latest update was done in August 2017. While there can be outliers in the data much of that is mitigated by the number of submissions. For more information see The Genetic Genealogist at https://thegeneticgenealogist.com/2017/08/26/august-2017-update-to-the-shared-cm-project ↩
- yDNA is the DNA that is passed down from father to son. ↩
- GEDMatch, “One-to-one DNA Comparison,” database V2.1.21(c), GEDMatch (http://www.gedmatch.com : accessed 18 April 2018), kit nos. withheld for privacy. ↩
- Ibid. ↩
- Genome Mate Pro is a program that allows you to compare and manage atDNA test results from multiple companies. It can be found here: https://www.getgmp.com/ ↩
- MyHeritage, Match comparison, MyHeritage (http://myheritage.com : accessed 22 April 2018). ↩
- The reported stop positions from MyHeritage are: Dad to Angus 225.7. Dad to Fiona 223.9. FTDNA shows the stop point for Dad to Angus as 223.5. ↩
- In very simple terms imputation is where the tested DNA is “expanded by inferring results.” Roberta Estes, “Imputation Matching Comparison,” DNAeXplained, 4 October 2017 (https://dna-explained.com/category/imputation/ : accessed 30 April 2018) ↩
- International Society of Genetic Genealogy Wiki (https://isogg.org/wiki), “Identical by descent,” rev. 17:29, 29 December 2017. LINK https://isogg.org/wiki/Identical_by_descent ↩