AncestryDNA has a new utility called ThruLines. The general idea is that they look at DNA matches and patch together trees on those DNA matches to the people who match. Sounds like a good idea, even though things could go wrong. The things that could go wrong are shaky trees and/or shaky DNA matches. However, let’s take a look.
Frazer ThruLines – An Overview
I have done a lot of work with Frazer relatives using Gedmatch. However, this new tool is quite interesting. When Ancestry comes up with a new tool, people take notice due to the millions of people who have tested their DNA there. I have access to ThruLines for some people in the Frazer DNA Project. Those people are:
- Myself and four of my siblings – I am on at least the Philip and Richard Frazer Lines – both from the main Archibald Frazer. Then I most likely descend from the other main James Line through Margaret Frazer and William McMaster.
- Gladys – She is on the Archibald Main Line also. She descends from the Philip, Richard and Archibald/Stinson branches of that Line.
- Jane – She is also on the Richard and Archibald/Stinson Line. Jane’s Archibald/Stinson Line seems to be better verified by genealogy.
- Michael – He descends from the same lines that Jane does.
- Richard – I believe that Richard is from the Philip Line. However, there are different degrees of uncertainty on many of these lines due to how far the matches go back and lack of paper trails.
- Joanna – She is on the James Main Line and comes down through James son Archibald born 1751
- Beverly – Beverly is on the same Main James Line but descends from James’ son Michael born 1764.
The DNA Family Trees
I have built some trees of Frazer descendants who have had their DNA tested. These trees have generally included those who have uploaded their DNA to Gedmatch. Gedmatch has made it easier to analyze the detailed DNA as AncestryDNA does not supply that detailed chromosome information.
The Archibald Main Line
Here is an overview of the Archibald Main Line:
The three brothers near the top are Philip, Richard and Archibald. Cousin Richard is in the green branch on the left. My siblings and I are in the blue branch under Philip, but also belong under the Richard Line due to a cousin marriage. Jane in the Army green line was added to the Richard Line due to all her matches there and seems to fit in. Gladys is actually the sister of the person circled and is in all three brother Branches. Michael is in the Richard Line and an abbreviated rendition of Michael is shown in the Archibald/Stinson line on the right above. David is included in purple due to a Frazer in his ancestry and DNA matches. He would be one of those branches with higher uncertainty.
The ones that I have circled are just the ones who I have access to their ThruLines. Many more Frazer’s have tested at AncestryDNA.
The James Main Line
I have access to only Joanna and Beverly on this Line, though many others have tested at AncestryDNA.
Below I cut off the left-hand side of the James Branch to save room:
I have also put my family in green where they may match up with the James Main Line. It is possible that the green branch should be under the Michael Branch.
James Line ThruLines
I’ll start with the James Line Branch as I have access to only two people in that Line. Joanna has already started doing her own ThruLine analysis on the James Line.
Starting slowly, we see Joanna’s connection with her dad:
This is the obvious stuff, but good research starts in the known present and works back to the unknown past. Going up a generation to Joanna’s grandfather doesn’t add any more matches. However, at Joanna’s great-grandfather level, there are some new matches:
Joanna’s grandfather has a brother William Fitzgerald Frazer who has 7 descendants who match Joanna by DNA.
Joanna matches four descendants of Minnie and three of Walter.
Let’s look at Joanna’s matches to Minnie’s descendants:
One thing I notice that is odd is that Joanna matches Steve by one more DNA segment than Kay. I wonder why?
What Are the Odds?
Next, all the match levels have probabilities associated with them when I click on them. Let’s start with the two second cousins, once removed. They are Chris and Kay.
Chris matches Joanna by 18 cM. Ancestry thinks that has a low probability of being right:
Rather than show each chart, I’ve summarized Joanna’s four matches on the Minnie Branch:
Steve’s likelihood of being a good match is much better than his mom Kay’s. 22% seems low, but it was the second highest category with 25% being the highest in this situation.
Joanna and Walter James Frazer
Here the relationships are a little closer but the DNA matches are much, much higher.
I’m not sure what to make of this. This seems to say that Joanna’s match with the Minnie line is more unlikely and that her matches with the Walter James Line are more likely.
Joanna’s Next Ancestor: Archibald Frazer born 1792
Here Joanna has only one match to give evidence to their common ancestor:
Remember above that Chris had a match of 18 cM but he was supposed to be a second cousin, once removed. WG is a 3rd cousin once removed. Chris had a bad likelihood score of <1%. Let’s see what WG gets:
Next Up: Archibald Frazer Born 1751
[Note: The genealogy in this Thruline is not correct. See my correction in my next Blog:]
Joanna’s in luck as Archibald 1792 had a brother:
That brother was some sort of hot-shot based on the size of his tomb stone. By the way, the 11 DNA matches shown above Joanna are the ones we already looked at.
CK and DK have low level matches with Joanna. We’ll put them into my spreadsheet:
When Joanna went back a generation, her percentages went up.
It’s Too Late To Turn Back: James Frazer Born 1720
Joanna has this James as James Frazer of Clerragh:
We are fortunate that Archibald born 1751 had a brother Michael:
Joanna is getting more DNA action on the Michael Line. Here there are five matches. There are two I haven’t opened up yet under Patrick Henry Frazer.
Joanna and Mollie
First off, Mollie and I share the same ancestors of Margaret Frazer and William McMaster, so I find that interesting. The other interesting thing is that Mollie and Joanna are shown as fifth cousins, once removed. That goes off the charts:
Ancestry is giving percentages for up to 5th cousin, but after that, they don’t even want to make a guess.
Joanna and the Michael Line
Here are the numbers for all of Joanna’s ThruLines:
I added in a column for cM. This to point out that Joanna’s matches on the James Frazer line were as high or higher as that with the Archibald Frazer line which is a generation closer. The DNA would suggest, apart from the genealogy, that Joanna could as easily or more easily descend from the Michael Frazer Branch than the Archibald 1751 Branch. Of course, other factors may come into play such as the randomness of DNA inheritance and matching on collateral lines.
Another interesting thing is Mollie’s ancestry. She is said to descend from Margaret Frazer which I do also. I had thought that Margaret descended from the Michael Line, but the more I look at the DNA, it makes me think that Margaret might descend from the Archibald 1751 Branch where Joanna has herself. If Mollie and Joanna both descend from the Archibald 1751 Branch, then that could explain the large match that Joanna and Mollie have.
Joanna has no additional matches up to the next generation to Archibald born about 1690. He was the father of Archibald and James Frazer – the two Main Lines.
Summary and Conclusions
- I had meant to go through all the Frazer descedants who I had access to at AncestryDNA and look at their ThruLines. That would take too long, so I’ll go through them at a slower pace.
- I started looking at Joanna’s ThruLines. Joanna’s matches showed some irregularities, but I would like to look at other Frazer descendant matches to see how they compare.
- Joanna had matches at the 2nd, 3rd, 4th and 5th cousin levels leading up to James Frazer of Clerragh. [Note: the 4th cousin match was not correct. See my correction in the following Blog.]
- Joanna had a larger than expected match to Mollie who shows the same mysterious Margaret Frazer ancestor that I have.
- I am only looking at the DNA here. The genealogy has to be taken into consideration.
- I am thinking that in the Archibald Branch with all the Frazer cousin matches, that ThruLine analysis could be complicated.
- I would like to look at Beverly from the James Line next.