More Recent Comments

Friday, December 28, 2018

On the accuracy of Ancestry.com DNA predictions

I'm very impressed with the DNA test administered by Ancestry.com. They report that I have over 600 fourth cousins or closer but I have confirmed some even more distant relationships. See below for the most distant relationship that the DNA tests reveal.

In the vast majority of cases the people who share DNA markers with me have no family tree that's on Ancestry.com so it's impossible to say for sure whether we are related. There are often clues based on who else shares our haplotypes but unless the person reveals their name and some of their ancestors that's all I can do. I usually contact those people who could hep me sort out some unknown relationships but I rarely get a reply.

The ones who already have a tree are much easier because then I have a list of names I can search and, furthermore, they are much more likely to correspond. We've found some of our common ancestors by comparing notes but in many cases we can't make the connection because one of us is missing some key links. The most interesting discovery was a 3rd cousin relationship from someone whose parent had been adopted but who didn't know their biological parents.

Some people have detailed family trees so it's possible to trace the exact connection and the results confirm the DNA prediction. There are no false positives so far but that's not surprising since it would be very difficult to prove that two putative DNA matches are NOT related. Each of the subjects would have to have a very detailed and accurate family tree extending back more than 7 generations and it would have to include all ancestors plus all siblings and their descendants.

The most distant connection so far is a seventh cousin once removed. We are direct descendants of two people who lived more than three hundred years ago! Daniel Robbins/Robinson (1627-1714) was born in Blair Atholl, Perthshire, Scotland. He was a Scottish soldier fighting on the side of King Charles I in the English Civil War when he was captured and sent to the colonies (Connecticut) as an indentured servant. There he met and married Hope Potter (1641-1687) and they moved to Woodbridge, New Jersey when he had served his time.


A seventh cousin and I share a French Canadian and a Dutch ancestor from the former colony in New York—Peter Montras (1715-1790) and Emmetje Anderzon (1717-1790). Peter was born in Phillipsburgh (Tarrytown), New York and he was baptized in the Dutch Reform church in Sleepy Hollow [Sleepy Hollow]. His parents are buried in the sleepy hollow cemetery (see photo). Emmetje Anderzon was born in New York (formerly New Amsterdam).


There are lots of fifth cousins. Here's an example of a connection to the Fraser clan in Scotland. John Fraser (1754-1807) was born in Inverness, Scotland just eight years after the Battle of Culloden. His wife Isabella Mackay (1762-1856) was also born in Inverness. They emigrated to Canada in 1803 with their five children.


So far I've established 25 direct connections to my DNA relatives and there are another 25 or so where we know our common ancestor but not the details. There are three problem groups in my tree but, unfortunately, the DNA results haven't helped sort them out (yet).


8 comments :

Jeffrey Shallit said...

My great-grandmother Emma Dean was adopted, so we never knew anything about her parents. But when ancestry DNA started showing I was closely related to people named "Dion", I put 2 and 2 together and figured out that her real birth name must have been Emma Dion. With more work I was definitely able to establish the relationship.

Jeffrey Shallit said...

Also, I don't know if you noticed, but Ancestry DNA recently modified their information so that they now provide shared cM info.

Larry Moran said...

Yes, I noticed. I'm not sure what to do with that information.

The biggest complaint that I have right now is that Ancestry only shows you the names of shared direct ancestors. I wish they would show those plus the shared names of anyone else in our trees.

Tom Mueller said...

https://www.livescience.com/63997-dna-ancestry-test-results-explained.html

Tom Mueller said...

Genealogy is becoming interesting and could spook potential criminals



https://www.haaretz.com/us-news/.premium.MAGAZINE-to-solve-cold-cases-all-it-takes-is-dna-a-genealogy-site-and-high-speed-internet-1.6657176

https://www.gedmatch.com/login1.php

EJ Blom said...

Hi Laurence, I created an interesting analysis for Ancestry testers (as well as for 23andme and FTDNA). It performs a clustering on the shared matches profiles. Here is an example: https://www.geneticaffairs.com/auto_cluster/examples/autocluster_40cM-400cM.html Usually, it performs this analysis in a matter of minutes.

Paul C. Anagnostopoulos said...

I enjoyed having my DNA analyzed. I learned that my father had his analyzed about a year and half before he passed away but didn't tell me. I share about half of his DNA, which makes sense. Ancestry matched me with two second cousins I knew and one I did not. It turns out she moved to Florida when we were both very young. They matched me with two second cousins once removed. Then there are a pile of fourth and further cousins. I haven't managed to figured out how we are related yet.

Unfortunately, my father's parents were born and died in Greece, so there is no information about them.

Unknown said...

Daniel Robinson birth 1627 surprisingly showed up in our line. We had 2 males tested with FamilyTree DNA Y only and found 1st male at 37 marker at a distance of 0 and the 2nd male at 67 marker at a distance of 1. We were all raised to believe that our last name was Barntd/Barndt and have put 20 plus years and 5 people at various times researching and the DNA showed no Barntd's except from my Gr Gr Grandfather down through. I called FamilyTree DNA and they said that the common ancestor was within 4 generations. My gr gr grandfather Barntd was born in 1815. We figure that our gr gr grandfather is the fourth generation from the 2 males tested. The information on Daniel is limited at best and I have to rely on what has already been published. You are the first that I have seen that has made many proven connections. My gr gr grandfather stated in a census record that his mother was born in New Jersey and his Father Virginia. Now that I know about Daniel,I think he may have had it reversed.I am basically at a loss at this stage of the research and feel like I am starting over. Anything that you can share or a direction would be much appreciated. Thank You and I assume we are some sort of cousins.