Note: This is the first in a series of posts telling this tale as it unfolds. Since the case involves living people, identities will not be disclosed unless and until all parties have agreed to publication. You may have guessed this by the names I’m using, but just in case…
For several years now, I have been working off and on on a brick wall project for a client to try and identify the parents of her 4th great grandfather. I had suggested DNA testing as a possible tool to help Esmerelda overcome gaps in the paper trail. She finally found a male cousin, Sam, to participate in yDNA testing, but no helpful matches appeared. After a year or two, she asked to have Sam’s kit upgraded to an autosomal test. When the results came back, we expected that Sam and Esmerelda’s tests would match. They didn’t. The chart below shows their relationship: half-first cousins once removed.
Half-first cousins once removed sounds a little diluted, and with only two kits to compare, it was certainly possible that one or both of them simply hadn’t inherited enough DNA to show up as a match to each other. It did raise the question whether or not they were actually related by blood. But again, with no more data, there was no way to know for sure. The paper trail was clear and convincing. There was no reason to question the relationship
Two years passed before another descendant from this family also tested, and both results were uploaded to GedMatch.com. The results were surprising. The two kits from other family members matched each other, but Esmerelda did not match either one.
Digging deeper into what the half cousin relationship should look like, it was clear that they should have matched. On average, that relationship would be expected to have about 212 cM in common.1 These kits had zero, zip, nada. The other two kits shared the amount expected with each other. Even if matching segments shorter than the recommended 7 cM were included in the comparisons between Esmerelda and the other two kits, it didn’t bring the total matching cM anywhere near the expected level of shared DNA. What the data suggested was mind-boggling: Esmerelda was not related to at least some of the people that until yesterday, she had known as her family her entire life. The question was where? There were multiple possibilities.
As of this writing, Esmerelda is contacting her closest living relatives, her first cousins, to see if they will test to help narrow down the possibilities. The DNA data from GedMatch pointed out the problem, but didn’t provide any theories as to its solution. That would require more data and research.
Next installment: Unexplained matches on Ancestry.com reveal a stunning possibility
- International Society of Genetic Genealogy Wiki, “Autosomal DNA Statistics,” http://isogg.org/wiki/Autosomal_DNA_statistics. [↩]