DNA testing for the Blennerhassett problem

The "Holy Grail" of my family tree was the Blennerhassett problem, the problem of figuring out how the Cashel family (our family) descends from the Blennerhassett family.

After years of gathering paper evidence, I turned to DNA.
In 2020, DNA finally proved we are descended from the Blennerhassett Baronets branch.

Stats: DNA samples so far. one-to-one DNA comparisons.

Note: In fact we are not checking some one-to-one comparisons, such as DAU versus LETITIA, since there is no point.
But we could do them later if we change our mind.

The general plan of DNA testing for the Blennerhassett problem

Gather the DNA of as many descendants of George Cashel as possible.
Target a specific branch of the Blennerhassetts where there is at least some paper evidence that they might be the branch we are related to.
Gather the DNA of as many people in that branch as possible.
Get their DNA on Gedmatch (where anyone can compare DNA with anyone).
Do one-to-one Autosomal DNA comparison of everyone with everyone.
Build up a map.
See what the map says.

The groups

We divide our samples into groups.

CASH - Our family

CASH are known descendants of George Cashel (born 1807).

DAU - Blennerhassett Baronets family

(The theory, based on some paper evidence, is that our ancestor comes from the Blennerhassett Baronets family. We named this group "DAU" since the family story is our ancestor was a Blennerhassett daughter. But this may not be true. We could come from a Blennerhassett son in this branch. So DAU is not a good name. We will change the name at some point.)

BART are known descendants of Sir Rowland Blennerhassett, 1st Baronet (born 1740 or 1741) and Millicent Agnes Yielding.
ROB are known descendants of Robert Blennerhassett and Frances Yielding.
YIELD are known descendants of Richard Yielding (died 1770) and his wife Millicent.
BUTLER are descendants of Thomas Butler and Tobias Butler. After some work, we have basically proved these are part of YIELD. For the proof see those two links.

LETITIA - Letitia Blennerhassett branch

(The theory, based on some paper evidence, is that our ancestor is Letitia Blennerhassett of Tarbert.)

BL are known descendants of "Black Jack" Blennerhassett (born 1665) and Elizabeth Cross
ED are known descendants of Edward Blennerhassett of Rossbeigh (born 1705) and Mary Fitzgerald
REV are known descendants of Rev. John Blennerhassett (born 1732-1733) and Louisa Goddard
FITZ are known descendants of Capt. Edward Fitzgerald (died 1738) and Jane Leader
GOD are known descendants of Rev. Thomas Goddard (born 1674) and Ann Beale
CAPT are known descendants of Capt. Thomas Goddard (born 1674) and Mary Mullins
MUL are known descendants of William Mullins (born 1691) and Mary Rowan

DISC - Disconnected Blennerhassett families

JOHN are known descendants of John Blennerhassett (wife born 1828), Catholic, of Tralee.
These were found by searching Ancestry DNA hits for my family for "Blennerhassett".
The DNA match may not actually come through the Blennerhassett line.
MCA are known descendants of Arthur Blennerhassett (married before 1841) whose descendants went to Massachusetts.
This is just a random disconnected Blennerhassett line that emailed me and turns out to have a DNA match.
The DNA match may not actually come through the Blennerhassett line.

DAU versus LETITIA - too distant

We could consider comparing DAU with LETITIA.
Most (but not all) of them have Blennerhassett ancestry, and descend from a common Blennerhassett tree.
But the common ancestor is too far back. You have to go back to Capt. Robert Blennerhassett (born around 1620s).
So we are not going to compare them. There is no point.

Guide to results pages

In the results pages, we compare individuals. e.g. "ED.1" is person no.1 from group ED.
People are only in these results if they have DNA on Gedmatch.
We consider segment matches in centiMorgans (cM).
Different sites use different minimum segment sizes. See autosomal DNA match thresholds.
9 cM is generally considered significant.
In the testing we compare many people who are known relations of each other. These matches show that:
- DNA testing really works. It roughly tracks the family tree. You tend to get smaller segments with more distant relations, larger segments with closer relations.
- You can get false negatives, even among people who we know are related. The more distant the cousins, the more likely to get false negatives.
- There is a lot of luck, with DNA being lost going down one branch, but not being lost going down a sibling branch.
- The more samples you have, in different branches, the better a picture you get.
With some comparisons we expect no match at all. Leave them in as a control. They show the conditions under which you can get a match by chance.
Black

is excluded. (Do not compare with self.)
Yellow

means we either cannot, or do not want to, do a comparison of these two people.
- Example of "cannot": A sample was on Gedmatch but is now gone.
- Example of "do not want to": We discovered that these two people match through a different line that is nothing to do with the Blennerhassett lines. So the match is not relevant here and it would be misleading to include it.

Disproved - Letitia Blennerhassett of Tarbert

The first DNA project I did was to look at the family of Letitia Blennerhassett of Tarbert.

From 2009 on, I constructed a theory from the paper evidence that our ancestor might be Letitia Blennerhassett of Tarbert.
See Theory of our descent from Letitia Blennerhassett of Tarbert.
After years of looking at paper evidence, I started a DNA project in 2019. This project aimed to DNA test our family and Letitia's family.
To my surprise, the DNA said that the theory is not true.
- DNA testing - CASH v. LETITIA

Proved - inside Blennerhassett and inside Cashel

The Letitia DNA project was not a waste of time, though, because the testing proved a large number of lines inside the known trees, and even proved new lines.

Proved - Blennerhassett Baronets

So, in 2020, assuming that we were connected to some Blennerhassetts, I looked for a "Plan B" branch of Blennerhassett to DNA test.
I soon fixated on the Blennerhassett Baronets branch.
So I started a project in 2020 to DNA test our family and the Baronets family.
And immediately, on the very first sample in 2020, DNA proved the two families are related. The Baronets theory is true.
After 11 years chasing the wrong line, once I abandoned it I immediately discovered the right line.
- DNA testing - CASH v. DAU

Unproved - other Blennerhassett lines

As I gathered people to DNA test, I also gathered some disconnected Blennerhassett lines.
It remains unclear how these connect to anybody else.
- DNA testing - DISC v. everyone