Saturday, March 22, 2014

Identifying descendants in Gedcoms to Compare your DNA to

In my last post I talked about looking for the parents of my ancestor, Mary Polly Allen. When I looked her up in, many online trees had her parents as Colonel Charles Allen and Ann Venable from Virginia. However, when I looked through the online gedcoms that were uploaded for each of my grandfather's matches on FamilyTreeDNA who had Allen in their list of surnames (there were 40 of them), I did not find a single one of them mentioning this ancestral couple. Since I struck out finding Charles and Ann in this list of matches, I decided to cast a wider net, though not by using the same methodology. is a 3rd party website that allows you to upload your DNA results and then compare them to results that were uploaded from the other two testing companies. Since my family and I tested with ftDNA, we are able to compare our results with people who tested with and 23AndMe. I am also able to compare to other people who tested with ftDNA. Why would I want to do this? Because I might want to be looking at lower resolution matches that did not meet the minimum 7cM and 700 SNP threshold that is required by ftDNA to be considered a match. offers several tools by which to find and compare matches; but it also offers some unique Gedcom utilities. One that I have just started using is the search feature. I can "Search All Gedcoms" for an ancestor and it will give me a list of all of the Gedmatch members who have submitted Gedcoms and their email addresses. I can then enter their email address into the Gedmatch User Information Lookup Utility (new) and it will tell me what their kit id is. Then I can run a one-to-one comparison between their kit and one of mine to see if there is a match.

In order to determine if Charles and Ann were my ancestors, I did a search of all Gedcoms for their names and it returned one match to a descendant of Charles Allen's sister Anne Allen. (The email was scrubbed for privacy).

Gedmatch Gedcom Charles Allen

I then entered the Gedcom owner's email address into the Gedmatch User Information Lookup Utility to get a list of their Gedmatch DNA kit ids to compare to:

Gedmatch User Lookup

(In this example, I entered my own email address and scrubbed some names for privacy).

Once I got a list of all of the Gedmatch Kit ids for Charles Allen's sister's descendant, I compared them one-to-one with my own family members who descend from Mary Polly Allen. Remember I am trying to determine if my ancestor, Mary Polly Allen (who married Abraham Huddleston), was the daughter of Charles Allen (who married Ann Venable) who was the brother of the Charles Allen mentioned in the Gedcom above. I can probably determine this if one of my family members is a match to one of this Gedcom owner's family members (he also managed multiple kits like me).

Gedmatch One-to-One Comparison between Allen Descendants

I entered the two Gedmatch IDs into the Kit No 1 and Kit No 2 boxes, and I entered a 1 into the Minimum segment cM size to be included in total. I like to add a 1 here so I can see the values for all chromosomes. Sometimes it's good to see if there is any matching DNA on a particular chromosome and you can do this by putting a 1 in this box.

I compared the oldest family members first: my grandfather and his father's aunt.

Gedmatch Comparison of Allen Descendants
My Grandfather vs Gedcom owner's Paternal Aunt

FamilyTreeDNA recommends to have a minimum of a 7.0 cM segment with 700 SNPs on that segment in order to be considered a match. The relationship between two people is determined by the total of these segments. In this comparison between my grandfather and his father's aunt, the largest segment is only 2.7cM (987 SNPs). This is nowhere close to being a match.

Gedmatch doesn't have such strict requirements. I believe they only require a 5.0 cM segment in order to be considered a match. When I compared my mother to the Gedcom owner's paternal aunt, there is the possibility of being a match:

Gedmatch Comparison of Allen Descendants-1
My Mother vs Gedcom owner's Paternal Aunt

On Chromosome 8, there is a 5.7 cM segment with 1,232 SNPs of shared DNA between my mother and the Gedcom owner's paternal aunt. There is also a total of 20.6 cM of shared DNA between them. Gedmatch has estimated that the MRCA or Most Recent Common Ancestor was 6.7 generations back. These are just estimates, so they are not exact. If you count my mother as generation no. 1, then James Allen, who is supposedly Charles Allen's father and the father of the Gedcom owner's ancestor, is the 10th generation ancestor from my mother. It is possible that this match with an estimated MRCA of 6.7 is from a different common ancestor all together.

10 generations is pretty far back. In fact, it is completely possible that this DNA test might not even pick up matches to an ancestor this far back! This test is only guaranteed to work back to 5 generations.


These inconclusive DNA results between my family and that of James Allen's family (the Gedcom owner) leads me to believe that my Mary Polly Allen probably did NOT descend from Colonel Charles Allen who was married to Ann Venable. Even though there is a possible match between his paternal aunt and my mother, there are no other matches between any other his other family members and mine, leading me to conclude we are not descended from the same common ancestor.

However, it is possible that this ancestral couple - Colonel Charles Allen and Ann Venable - is too far back to be picked up by the autosomal test. But I really don't believe that is true. Mary Polly Allen is only my grandfather's 7th generation back (if you include my grandfather as generation no. 1). See chart below. I have found common ancestors with 8th and 9th cousins with several of my matches, so I know it's possible to do.

7 generation chart to Mary Polly Allen

But DNA is fickle. Actually, not really. But there are so many factors to consider and possibilities to ponder. My main goal of this post was to show how you can try to verify or eliminate your questionable ancestors by identifying descendants in Gedcoms to compare your DNA to.


  1. I wouldn't discard the possibility that your grandfather and the great aunt are both descended from Charles Allen's parents. If I'm doing the math right, these 8th cousins have a 7 in 10 million chance of sharing a segment of 5 cM. So the chances are great that 8th cousins don't share any DNA inherited from Charles Allen's parents. If you have enough 8th cousins, though, and you happen to match one - chances are that piece is IBD. See

    1. Thanks I will definitely keep that in mind. Working with autosomal results is not always as easy as it may seem! Thanks also for the link.