How far back do the autosomal DNA results go back? & DNA Connection [closed]

Question

I know we've discussed this before, and I was under the impression that the autosomal results would appear on the profiles of up to 4 or 5 ancestors back, but with my grandfather's profile, his autosomal results are only showing up on his profile, Larry Godwin (Godwin-648), his mother, Nova Martin, and his grandmother, Dora King. I would like to go back to at least his 2nd great grandfather in order to create a DNA connection with another House cousin. (Actually it looks their common ancestor goes back one more generation to his 3rd great-grandfather).

Comments

I see now that it is 6, but could be 3 up and 3 down. So with my grandfather's mother's side, it is 3 up only; but with his father's side, it is 6 up. So why the difference?

---

I haven't done the DNA test yet for various reasons.  If this connection isn't 100% provable, a person would still be searching for that link.  I'm not sure if I am making my point clear, but that is how I am seeing it.  Please correct me if I am wrong.

---

Hi Ginger,

It should be six steps in any direction.

I do see connections on Lawrence's mother side out to five or six steps, e.g. http://www.wikitree.com/wiki/House-1975

http://www.wikitree.com/index.php?title=Special:DNATests&u=7013140&id=6

http://www.wikitree.com/index.php?title=Special:Relationship&action=calculate&person1_name=Godwin-648&person2_name=House-1975

What's a profile (WikiTree ID) that it should be attaching to that it's not?

Chris

---

Hi Chris, thanks for responding. When I merged my Hollingsworth House with my cousin Earl's profile of Hollingsworth House (House-1973), that then attached us to Jacob House, etc. That was when the DNA note showed up past Larry Godwin's grandmother, Dora King . So,  it was either the merge or the "it-takes-24-hours-for-our-database-to-referesh" comment from the WikiTree team that did the trick - I had just added to the family that day.

---

I recently triangulated a match going back to my 6th great grandparents, 8 generations, at 22.8 Centimorgan on one cousin and 16.11 with the other.   The shared portion of the segment was 11.79 Centimorgan and the cousins shared another 12 Centimorgan match that I do not match on.  One of them had this set of ancestors in their tree twice.   I was surprised at the strength of this match and did a due diligence to confirm,  but since it is a well documented DAR ancestor the trees were solid.

I would be interested to know how far back others may have been able to do confirmed triangulations, what is the realistic edge of the envelope, given the understanding that actually finding the rare cousins who will actually match at this distance is a great challenge in itself.

---

Lynn, this query to which you attached your questions is almost four years old and was dealing with a different matter.  I would suggest you post your comment/question as a new G2G post...

---

I thought I might sneak the stupid question in this way.  LOL

---

Lynn, it isn't a stupid question.  I just think you will get more responses if you start a new thread.  Autosomal can go back up to ten generations.  Generally you won't be able to determine matches back that many, though.  But it does happen.

---

Thanks.   I keep seeing trees with no apparent link other than Robert "Potato Hole" Woodson and that's 10 generations.   At that distance there are so many empty positions on most trees that it adds yet another element of risk.  I was wondering if I was being silly to even make note of the matches.   I'm not sure I would even call it a valid match unless I had more than two cousins in three triangulation.  After I got 22 cM at 8 generations,  I started thinking I might be able to go father.

---

Lynn, I've been attempting to get a triangulation on a family that would be 7th cousins. Our assumed common ancestor would be 6th g grandparent I believe. I match 11.2 cm with two brother's & 8cm  with their great aunt. I'm told this is not a good triangulation & I have to find a different match not related to confirm. Very frustrating. I'm also told this is borderline distance for a triangulation.

---

Correct.  All parties of the triangle need to be at least 3C1R distant, I think.  You need to be able to rule out the possibility of matching the others on the same unidentified ancestor within a tolerable level of risk, type of thing, so people in the same family don't work.  I guess I got really lucky to find two of them, the chances of anyone matching at that distance being extremely low, but then there are a lot of descendants as you get farther out too.

Answer 1

I'm not sure why that is happening on Wikitree. As has been discussed the tests are good for about 5 generations. This link helps explain:

http://www.dna-testing-adviser.com/Autosomal-DNA-Testing.html

Comments

From my (admittedly limited) understanding, saying auDNA is good for about 5 generations is perhaps misleading. 5 or 6 generational steps would perhaps be more accurate.

From what I can find, third cousins (7 generational steps) are ~90% likely to have enough shared auDNA (about 0.781% commonality) to detect a relationship. A shared 3G Grandfather suggests a 4th cousin, possibly with some degrees of removal. 4th cousins are ~45% likely to have about 0.2% auDNA in common, while a relationship of 4th cousins twice removed is ~15% likely to share about 0.0488% auDNA.

*Edit: this assumes that only tests for the two individuals are available for direct comparison.

Answer 2

Hello Ginger,

We have about a 50% chance of having no auDNA matching with our 4th cousins, and 90% chance of having no matches with our 5th cousins, and a 98+% chance of no matches with 6+th cousins.  Most people have a huge number of 4th cousins (and higher).

GEDmatch and DNAGedcom and the testing labs tell us who our matches are.  Share the link to your ancestral tree in WikiTree with those cousins and invite them to upload their ancestral tree.  It would be a lot of wasted time to be contacting distant cousins out of the blue who you only have a remote chance of sharing an autosomal segment with.

Sincerely, Peter

Comments

Peter, I recently found out that I share a 16 cM segment, on chromosome 3, with my 6th cousin. I say "6th cousin" because that is what our documented genealogy says. Does this DNA result have any meaning in confirming, supporting, or refuting our genealogy?

 

Besides "6th cousin", Family Tree Maker also calls this relationship "Civil Degree XIV" and "Canon Degree 7". In other words, there is a 14 generation distance between me and my 6th cousin. 

---

Hello Rick,

We can share autosomal segments with some 6th cousins.  Is the segment you share with your match inherited from the ancestor (or one ancestor of the ancestral couple) you have identified, or could the segment be from another (more recent?) ancestor you have not identified?  This is why "segment triangulation" (as found in GEDmatch and DNAGedcom) is so important.  If a group of people all match each other on the same segment then you can then look at what is the common ancestry of most everyone in that group.  That should reveal which ancestor (or ancestral couple) the segment is really from.

I want to imagine that it will one day be possible for WikiTree to automatically identify shared ancestry between a group of people who all match each other on the same auDNA segment.

I want to add that it would be a big mistake for WikiTree to automatically show on our profile which of our 6th cousins have had an auDNA test because most of us will have far too many 6th cousins but we will not share any segments with most of them.

Sincerely, Peter

---

Rick,

16 cM is closer to the range expected of a fourth cousin, but expected values are simply mathematical assumptions based on 50% commonality at each generational step. The amount of commonality could actually be higher or lower at each step. Additionally, it assumes that there is only one 'blood relation'.

For one example, if a parent is an identical twin, that generational step is ~100% commonality rather than the assumed 50%. Similarly, a first cousin is expected to have 12.5% commonality in auDNA but a double first cousin, related through two blood lines is expected to share 25% auDNA. Combining these two ideas, If two sets of twins marry, the children of each couple could be mathematically expected to share 50% DNA with one another, the same as is expected of siblings.

---

I up voted this question not because I agree with it but because the question is very interesting to me. I don't know enough about DNA to have an informed opinion one way or the other, so I can neither agree nor disagree with the statements made therein.

In any case, thank you for posting.