Has Anyone Used Autosomal DNA Test to Identify Unknown Great-Grandparent?

+6 votes
543 views

Wondering if anybody in WikiTree community has used autosomal DNA testing to identify unknown great-grandparent?  My maternal grandmother (deceased) was born out of wedlock.  Only birth record available online identifies person as father.   However, an uncle said many, many years ago that the Son of man (listed on birth record), was actually her father.  Found no documentation evidence that identifies Son as the father.  Contacted town clerk at location my grandmother was born and town birth record matches online record.  Think only option of identifying father is DNA testing.  I recently took AncestryDNA autosomal test.  Close (cousin) match is elderly grandchild of Son, with possible DNA relationship of ‘Half 1st cousin 1x removed’. If accurate, points to Son as my grandmother’s father.  Only living descendants (oldest generation) of Son are (elderly) grandchildren.  Same for my grandmother.  Know I need to compare DNA with more direct descendants of Son (preferably grandchildren).  If anyone has any advice or experience with this type of genealogical puzzle, I’d be grateful if you could share your story.  Thank you.

in Genealogy Help by Dan FitzDaniel G2G Crew (940 points)

3 Answers

+7 votes
It sounds like you have a possible name and place, and you are just not sure exactly who the father was? As opposed to just -- there *must* be a father, but you nothing except the surname the kid was born with?

If it's more like the former, then I identified my mother's 2x-great-grandfather Van Meter with good certainty using her autosomal DNA test. I knew my great-grandmother's probably main ancestor with her surname,and I knew her father was John Van Meter. But I didn't know which man of that name, so I couldn't determine the line down to her.

I created a big spreadsheet with every match I could find who had a tree showing a Van Meter in the same general family (starting by searching the match list for people with Van Meters in their trees, including variant spellings, then looking at the shared matches with those matches to find a few more even). In many cases, I noted errors in their trees and had to fix them with reference to land and probate records, etc.

Once I had a couple dozen names or so with the correct lines identified, there was a pretty clear pattern with the closest matches mostly descending from one guy (named John!), the next most distant matches descending from his father, and the next most distant from *his* father. It was fun partly because I'd have a close match who seems to descend from the most distant ancestor, but then I'd find a mistake in their line and it would turn out they are a closer relative. That happened a few times. It was also fun because the man I identified as the 2x-great-grandfather was not someone I had looked at before -- he lived mostly in PA, and I was expecting NJ, so I hadn't really explored the PA records well.

There's an image of the spreadsheet on the profile of the man I identified as the 2x-great-grandfather:

https://www.wikitree.com/wiki/Van_Meter-861

I actually have a few more rows in my local copy of that sheet, so this reminds me to update the image.
by Barry Smith G2G6 Pilot (319k points)
edited by Barry Smith
Did the same thing and found my 3rd ggf.

Thank you Barry and Mike.

Barry - I'm going to use your 'algorithm'.  I'm creating a copy of your spreadsheet.  Appreciate your help.  Dan.

+6 votes
I can relate as I had a similiar experince with a great-grandmother.  My mother's tree has been accurate at the 95%+ level, so what I found was a bit of surprise.  I questioned by why my Xggm was born 6 years after her father died.  And, she had his last name.  I thought it bad dating one way or another.  So, I started examing the census records, and lo and behold the dating was correct.  

One census however showed her at three years of age and her mother and another man old enough to be the father.  Even better he had a name.  So, I added him to my Ancestry tree (yes, that is sacrilege on WikiTree) and his parents and grandparents. And then waited to see if there were any DNA matches. It only took several weeks and all of a sudden it started raining half-cousins.

I am currently working on a maternal Xggf that is a brick wall.  I had a male cousin on my mother's side take a full spectrum DNA test to see what might hit on autosoma, mtDNA and Y-DNA in that line.  Keeping my fingers crossed.  I just wish I didn't have to wait weeks for the Y-DNA results.

Ancestry has it's faults but it does have an advantage is you are statistically minded. When stumped to a general name search and see what pops up in terms of spouses and parents. It can be all over the place.  However, sometimes the majority grouping of like kind is correct, and other times it's like lemings over the cliff.  And that single outlier on the path less travelled is the key.  Amazing, one can see the patterns.
by Malcolm Holcombe G2G6 (7.3k points)
+5 votes

One useful tool, which will calculate the probabilities of different possible relations, is "What are the Odds". It's available on DNA Painter at here. You could put yourself and the cousin in, and test the hypotheses that your common ancestor is (1) the son and (2) the father.

by Harry Ide G2G6 Pilot (101k points)

Related questions

+19 votes
3 answers
+20 votes
8 answers
+6 votes
1 answer
+12 votes
1 answer
+17 votes
2 answers
+8 votes
2 answers
331 views asked Jul 8, 2019 in Genealogy Help by Rhonda Zimmerman G2G6 Pilot (234k points)

WikiTree  ~  About  ~  Help Help  ~  Search Person Search  ~  Surname:

disclaimer - terms - copyright

...