Have any second or third cousins taken DNA tests? Simplified DNA confirmation.

Question

Hi WikiTreers,

Have any of your second or third cousins taken DNA tests for genealogy?

If your close cousins are on WikiTree you'll know if they've taken a DNA test because it will appear on your profile. I recently had a second cousin and a third cousin appear on mine.

With third cousins or closer, it's easy to get started using the "Confirmed with DNA" relationship status indicators  -- potentially back to your great-great-grandparents.

Here are the newly-rewritten instructions: http://www.wikitree.com/wiki/DNA_Confirmation

And a personal example to follow: http://www.wikitree.com/wiki/DNA_Confirmation_Example

What do you think? Does these instructions make sense?

DNA Project Members and other experienced genetic genealogists: Any advice or suggestions for changes? (Andreas: I think you may have had some suggestions for changes on some of these help pages. Forgive me if those are still outstanding.)

Chris

Comments

Love it!  Really easy to read and follow, thanks :)

---

I'm confused about the X-chromosome Confirmation section. It says, 

If you and your match are both men there is a special opportunity for one-to-one confirmation if you match either on your X chromosomes.

If you match either (what?) on your X chromosomes? 

I would prefer a bit more explanation in this section. 

---

Hi, Chris:

I have been using the "Relationship to Me" feature ever since it first was available on WikiTree. Have found so many distant and near cousins, etc. and have communicated my findings with many of them. However, I had wondered where to go from here? Today, while perusing the G2G feature, I came across your wonderful question - yea! I will now follow through. My father is definitely related to me by blood and DNA, as well as grandparents and great grandparents, etc. Have wanted to change the unknown confidence to confidence for quite some time on many of my ancestors, now I can. Thanks so much ...

---

Hello Jeannette,

Do you have any third cousins or closer relatives in WikiTree who have had an auDNA test?

Sincerely,

Answer 1

Lesson learned is that you can't trust estimates like "your match could be a third cousin or closer" its all dependent if you have a big or small "genepool". 

I suggest add a comment box next to the DNA confirmed box so you can motivate the decision....

Picture above is my ICW from FTDNA FF test visualized at http://DNAgen.net   at the bottom you can see a lot of purple names that are all connected with each other which I think indicates that those people are all related to a small genepool and that I cant trust the estimates of our relationship to those peoples as it will be much higher than it should.....

E.g. user Bildtse-2  has her whole family tree in the south of Sweden ==> that "everyone" is probably related in one or more ways.... ==> you have a small gene pool the estimates will be much higher than they should....

A classic example is Ashkenazi Jews that have a really small genepool and when they test on FTDNA they get enormous amount of hits just because everyone is related to each other and the value they get is not related to the "distance" ..... see Endogamy

Another aspect is that companies like Ancestry and FTDNA are changing the algoritm for estimates all the time ==> what today is a thrd cousin maybe isn't tomorrow see below...

• Autosomal dna research challenging 
  "As a result of frequent intermarriage, a Family Finder cousin match may show a total value of centiMorgans composed from the combination of different lines. That is, they are a more distant cousin who is related in multiple way"
   
• Chromosome Pile-ups is something new that we believe makes estimates wrong
  "Illustrate the need that AncestryDNA had for redesigning their matching system to account for population wide shared chromosome regions"

Comments

Thanks for this, I have been worrying away at at the self same issue where descendants of [[Kingman-4 | Henry Kingman]] are coming up with autosomal matches far larger than the paper trail supports.

It increasingly looks like this is down to a large number of cousin marriages in what was, at the time, a relatively confined gene pool.

---

If you use FTDNA test http://DNAgen.net it could at least give an indication if more people are more related than others ;-) 

---

Please all keep in mind that ICW isn't triangulation and thus these people can be related on different parts of their family tree (eg on their paternal and maternal side at the same time).

So it doesn't always mean that this is a large group of DNA cousins that is all descendant from one common ancestor! ICW in most cases is on different chromosomes and triangulates only in way less than 50% (at least for me).

So what you see might be rather population wide.

As for chromosome pile-ups there are several reasons for it. Population wide is one, a common ancestor with a lot of children (usually with different wifes) and his children having themselves a lot of children is another. Lastly there might be still the chance for IBC (by chance) which means even though some people triangulate the usually very low density (defined as the number of connections / the number of possible connection between all people in the triangulated group) could be a consequence of a computer algorithm stitching a maternal and paternal segment together.

---

I'm sorry.  II am SO LOST with this DNA stuff.  I took DNA tests with FTDNA and I am told that I have Ashkenazi Jewish in my results.  I have not seen a graph like you have here.  I have red and read and re read, and I am still lost on what to do.  I am ready to just give up.  I've been at this for a year!

---

Hi Ken, Do you have any matches for either your yDNA test or your Family Finder test? Do you know how you're related to any of your matches? If your DNA matches someone else's, your next step is to compare their tree to yours to figure out which ancestor you have in common.

On your profile you haven't indicated that you've uploaded your Family Finder test results to GEDMatch.com. That would be a good next step. On FTDNA you can only see people you match who also tested at FTDNA, but if you upload your Family Finder test to GEDMatch, you can find people you match who tested at 23andMe and Ancestry (assuming they also uploaded their test results to GEDMatch.com).

Answer 2

Would it be possible to expand on how to get a TG group please?

In the following example I am using One-to-One 

ie Use Gedmatch triangulated matches and find overlapping segments. But how do you break it down to the segment needed for confirmation?

I believe I would take the highest start point and the lowest end point, but how would I get the centimorgans?  The below is a bad example because I would use the 1st kit's figure's in this instance, but let's say I had to use the start from #1 and the end from #2, how would I then calculate the cM's?

I believe this needs to be covered within WikiTree as a lot of people want to add their DNA but really have no clue what to do, the elderly don't use Facebook to learn etc.  The more detailed help we can offer the bigger chance we have of them adding the info.

Chr    Start Location    End Location    Centimorgans (cM)    SNPs

A889678 RL 
11    31,996,185    69,310,220    27.7    7,501

M224633 PL
11    26,318,150    74,893,069    37.6    9,545

F386656 MF
11    30,351,310    75,653,949    35.6    9,045

F369086 JM
11    26,332,682    77,451,574    40.1    10,233

Devil's advocate...

Comments

So are you wanting all the 22 Chr listed or a shortened version. Comparing 2 or 3 kits for triangulation--I'm still new to the DNA stuff, but willing to learn more

or something more like the Autosomal DNA comparison--would that work with total cMs--which you can compare like 3 or more DNA samples

Value shown is cM total of matching segments over minimum threshold.

Kit	name	A377136	A601929	A129628
A377136	*L.C.		562.5	421.6
A601929	*L.M.	562.5		1892.2
A129628	*R.R.C.	421.6	1892.2

---

Thank's but no, I was talking in relation to the original post about templates for DNA confirmation.  See here: http://www.wikitree.com/wiki/DNA_Confirmation#Autosomal_Triangulation

I was giving a question an example of how to get to that point but needed help in calculating the cM's.

---

Ok I went from above your question, the one from Chris, .sorry Rhonda.

I understood what you were asking now, it was late. So if you had at least 3 matches with 7cm or more overlapping segments from GED match (one to one), would there be enough room to make this show up looking nice in the sourcing section of the DNA confirmation section

---

Maybe just take a screen dump from gedmatch.... feels even easier ;-)


http://www.wikitree.com/photo/png/Salgo-1-2

Difficult/impossible to process further but easier to upload.....

---

@Lori: you example above is questionable for triangulation as two of the three people are actually to closely related (1892.2 cM).

In general I've read so far that every from 1st cousin once remove upwards should be handled with a question mark. That means it can be a TG but it also can be IBC (by chance).

Once you find other DNA cousins (which are 2nd cousins and further out) triangulating in the above group you have your proof that it's a valid triangulation!

---

Thank you

---

@Lori - please don't get my answer wrong, it might be a valid triangulation. Thing is, we don't know this for certain (meaning 100%) right now. With more matches or other TG's you find at the same position we will know.

Please do also read the link I posted for Jim Bartlett's blog about "walking the ancestry back". That will further help you to verify TG's (it's in another part of this answer I think, wrote too many answers tonight)

Answer 3

A small comment about the instructions. In the section about triangulation, it says that family finder does not let you see how your segments match. 

It does, but you need to use the Chromosome browser tool which will let you compare up to five named matches at one time and see exactly which segments of which chromosome you match on.

Comments

Thanks, Derrick! Care to write out the steps?

---

Hello Derrick,

Unless you are a FTDNA project admin with the participants in your project, then you can't see how your matches match each other.  You can only see how you match them. Family Finder testers have to use GEDmatch to see  if your matches match each other on the same segment or use GEDmatch's triangulation tool which is one of their paid Tier 1 utilities.

Sincerely, Peter

---

Chris

Please see Peter Robert's reply below. Peter is quite right, a project admin (e.g. me) can do this, but a project member cannot without getting the other matches to cross correlate their own results.

A "Duh!" moment for me when Peter pointed this out.

Answer 4

I have a question about more distant cousins and triangulation.

The case in point involves me, a second cousin and a fourth cousin. The paper trail defining these relationships is supported all the way through.

The atDNA tests that we have each done show a relationship of about the right size of match, but:

me and the fourth cousin share a big chunk on Chr 4 (17.8 cM) and a smaller chunk on Chr 1 (2.8 cM)

me and the 2nd cousin share several sizable chunks but the important ones in this case are that there is no overlap with the chunk on Chr 4 above, and on Chr 1 we share a chunk of 15.3 cM

the 2nd and the 4th cousins share a big chunk on Chr 1 (22.2). This chunk is immediately before the one that I share with my 2nd cousin and includes the segment that I share with our common 4th cousin.

On that basis we don't "triangulate" as we only share 2.8 cM on the same chromosome.

Question is, can this data be used in any way within the wikitree system to show a confirmed relationship?

Comments

Good question.

Derrick, you should ask this as a new G2G question, to make sure more people see it, and that the answer serves as a better reference for others who have the same question in the future.

I think the problem here is that you don't know exactly how you're related. You could be related to each other in different ways via different ancestors. But I really don't know what I'm talking about on this. :-)

---

There's been a discussion of this here

http://www.wikitree.com/g2g/236215/for-triangulation-how-large-does-matching-segment-need-to-be

The most you can use that small section, according to current standards, is as a pointer when doing other triangulations.  You need 7 cM for each tester on the same segment.

---

Thanks very much for that link Kathleen. The answers given there are very helpful.

---

Chris and Kathleen are correct. You need to find more DNA cousins that match into this group of three. Once you find a 4th and 5th person the chances for a true triangulation are increasing by a large manifold.

If you have matches and all of them aren't triangulating into those or you actually find two other TG's at the same spot (where each one in the 2 other TG's doesn't match with this original group) then it's a bad apple ;-)

Another method is described by Jim Bartlett in his blog which is called "walking the ancestry back". See his excellent blog article (and please read the rest as well, he's one of the most experienced in DNA genealogy) here: https://segmentology.org/2016/01/02/ca-and-mrca/

Answer 5

For xDNA, I don't think it makes a difference in the sex of the DNA Matches.

IMO, it should only be used to answer the question "which sex". If 3 DNA testers are documented to be full 4th cousins, then they would share the same 3rd great grandparents.  For 3rd cousins and closer, we can confirm both Most Recent Common Ancestors (MRCA) because the likelihood of predicting 3rd cousins or closer accurately, when 2 DNA testers share only one MRCA is remote.

This is why when 3 4th cousins share a single segment, that segment came from either their 3rd great grandfather OR their 3rd great grandmother. In the case of 4th cousins, also sharing an xDNA segment may be used to confirm the DNA of the 3rd great grandmother if either of the 2nd great-grandparents, children of the 3rd great grandmother, are male.  

It doesn't matter if the DNA testers are male or female.

Comments

Just to be sure we are using the same terms, a 'match' indicates that a pair of DNA testers are believed to be related within a genealogical timeframe.

auDNA is used to determine genetic distance and xDNA may or may not be part of this formula. 

23andme for example...

For half-IBD shared segments, the minimum thresholds are::

X (male vs male): 200 SNPs, 1 cM
X (male vs female): 600 SNPs, 6 cM
X (female vs female): 1200 SNPs, 6 cM

 http://isogg.org/wiki/DNA_Relatives 

The minimum threshold must be met and once it is met, the gender of the two DNA testers follow the same rules for xDNA.  male DNA tester(s) follow the rules for males and female DNA testers follow the rules for females.  

If 2 males share an xDNA segment and meet the minimum requirements, they should be treated the same as a female and male/female pair who meet their respective minimum requirements.

 

---

Great that you clarified the different minimum thresholds for X-DNA, Ken!

The X-DNA is often a problem for triangulation, especially when both testers are female (meaning you get loads of matches, eg the majority of my partners matches are all on X-DNA). Unfortunately they are very small and thus not worth to further follow them up (meaning too far back in time).

---

Andreas - The 3 major DNA services don't list "Matches" based on the xDNA. That is because you can't determine genetic distance based on the xDNA length.

Gedmatch xDNA support is pretty basic and treats males and females the same.  I see xDNA segments in my gedmatch list that are not reported in the DNA Service because the threshold is too high when a male is involved, and comparisons involving 2 females are reported because the threshold is too low.

The default is still 7 cM and 700 SNPs.when you do a X DNA compare. 

I took a look at your gedmatch list and see that M419335  is a match that reports no xDNA segments. Given you are both on 23andme, I changed the defaults to 1cM and 200 SNP's.  The gedmatch result shows that you share a 1.3 cM with  290 SNPs. Ths tells me that you both are related to each other on your mothers sides. :)

I don't see the difference between triangulation that involves 3 males who have at least 1cm with at least 200 SNP's and 3 females that share a 6cm or some combination that meet the minumum requireents.

---

Sorry Ken, I don't understand this remark of yours: "The 3 major DNA services don't list "Matches" based on the xDNA."

23andme shows X-DNA matches to me.

As for your comment about GEDmatch I think their purpose is a different one than the DTC's. GEDmatch was and is always about the freedom of entering your own parameters. As such, whatever you enter will be used to determine your matches that are shown as GEDmatch results. Surely you can enter something "wrong" in there and GEDmatch will still show it.

But it's essentials as currently no one else offers this flexibility and for that I think we're all thankful to the guys behind GEDmatch. Yes, they could offer to automatically set the "correct" (by current standards) thresholds for X-DNA 1-to-1 comparison instead of going with the default 700 SNP and 7cM but hey, it's a hobby project by two guys with a normal 9 to 5 job (as they told me).

You're correct that James, my mum, myself and a group of other people all triangulate with each other. It's an excellent remark to check the X-DNA for male-to-male matches as a valid match indeed immediately eliminates the whole paternal side for both matches, a great step forward in further narrowing down the MRCA!

---

It sometimes takes me a few attempts...:)

I don't want to stray too far from my original point that that it should not make a difference if the DNA testers are male or female.  The differences in sexes are addressed in having different requirements as to what is and is not a valid xDNA segment.

The documentation indicates that "If you and your match are both men there is a special opportunity for one-to-one confirmation if you match either on your X chromosomes. Since men have only one X chromosome, no triangulation is necessary."

The fact that the DNA testers are both men does not tell us anything about the possibility of one-to-one confirmation".  If the female common ancestor only had daughters, you can not do a one-to-one confirmation based on xDNA, regardless of sex. If you have 2 DNA testers that are female and the common ancestor(s) only had sons, you could.

From what I read and is also stated elsewhere, is that an xDNA segment alone, without reference to auDNA is sufficient to confirm one-to-one confirmations. My point is that the DNA Services do not consider the xDNA results reliable enough to use in the calculation which determines if a person is a match.

It is true that once the determination that a person is a match, the xDNA comparison results are shown for the 3 DNA Services.

---

Hello Ken,

You said "If the female common ancestor only had daughters, you can not do a one-to-one confirmation based on xDNA, regardless of sex." In this case triangulation is necessary.

If those two daughers each had a son then you can do a one-to-one confirmation between those two sons using their xDNA.  No triangulation is necessary.  

Sincerely, Peter

---

Peter, I am confused as to the meaning of one-to-one.

I take this to mean that 2 DNA testers can confirm their relationship up to and including at least one common ancestor, similar to how yDNA and mtDNA work.

If 2 DNA testers are confident of the connection as full 4th cousins via a paper trail, this means that they share the same 3rd great grandparents. In my mother's case, this may mean that they were born around 1750.  

If these DNA testers, born around 1940, are related to each other via 2 daughters of these 3rd great grandparents, there is no way to get a one-to-one confirmation between any two DNA testers, male or female.  These daughters would be born around 1775 and each of their sons would have been born around 1800, so we can't use their xDNA.

Even if you could use xDNA of son of the 1st daughter and the xDNA of the son of the 2nd Daughter, you still can't determine if the xDNA came from their grandfather or their grandmother

Answer 6

I think the following is technically incorrect.

"You will need at least three test matches (you and two or more distant cousins) who all match on the same segment of DNA to confidently confirm which ancestor(s) you have in common."

Each shared segment can only have come from 1 person.  If the documented tree indicates the testers share 2 MRCA's, you can not confirm which of these ancestors you have in common. sharing xDNA is a possible exception.

 

 

Comments

Correct.  If three or more cousins all share two different ancestral couples then you don't know which of the four ancestors contributed the shared segment.

---

Peter, if three or more cousins all share the same most recent ancestral couple, you still don't know which ancestor contributed the shared segment. A single segment can only come from 1 person and you don't know which Ancestor.  The text indicates that this 1 segment can confidently confirm which.  It also indicates that this 1 shared segment could possibly come from more than 1 person by stating that you can "confidently confirm which ancestor(s) you have in common."

---

My two cents worth. This one segments is only coming from one CA. -> true

The problem is that we don't know who of that CA couple is responsible for that ancestral segment (exception for X-DNA, which the paternal line can't give to his descendants)!

The only way we can identify the CA for atDNA (all the non sex chromosomes) is when the CA becomes the MRCA and we have to search for the CA further upstream in the family tree (where again we don't know who in that couple gave the segment - unless it's X-DNA).

As for that quote "If the documented tree indicates the testers share 2 MRCA's, you can not confirm which of these ancestors you have in common." the two MRCA's must be connected by the CA. If you can't find a connection between the two MRCA's (because of not enough depth in the family tree) then it's correct that you can't identify which of the couple was giving the ancestral segment down - exception for X-DNA again).

I think it should be formulated like this. A picture would probably help to illustrate the problem better. One showing how the two MRCA's connect to one CA and one where the family tree upwards ends and doesn't connect (so you have two separate trees).

Is that what you (Peter and Ken) agree with?

Answer 7

I have a 2nd cousin 1x removed who has tested with Ancestry DNA, our estimated match is predicted to be "distant" cousin. We also have a two DNA Circles in which we share the same grandparents, my 3rd Great-Grandparents are her 2nd Great-Grandparents. In this case, am I able to check confirmed with DNA?

Comments

If this person is your second cousin 1x removed, your DNA match should be a whole lot closer than "distant cousin." You are definitely not "confirmed with DNA."

I suggest that you both upload your data to Gedmatch so that you can see where you match, instead of trying to figure out Ancestry's hocus-pocus.

The lack of a closer match with this cousin suggests that you may be on a path to some surprising new discoveries resulting from your DNA. Keep your seat belts fastened!

---

You would be 3rd cousins 1x removed, not 2nd (3rd ggp and 2nd ggp).  For this distance, you need to, as Ellen mentioned, both upload your results to Gedmatch and compare DNA.  Then you need to find another person that descends from your 4th great-grandparents (your most recent common ancestor with your known cousin) and that shares the same segment (over 7cM) with both of you, i.e. you all share a common segment.  Then you have a triangulated group and can confirm the relationship.

---

I went back and looked at the DNA circles and realized that she is no longer in the list of profiles in the two DNA circles for grand parents I referenced. It seems like what ever "update" Ancestry did altered a lot my matches. Thanks for the suggestion about GedMatch.

Answer 8

Chris, 

My 2nd cousin once removed (female) and I (male) match on a GEDmatch X-DNA Comparison with the largest segment matching 108.0 cM. 

I followed the Help:DNA Confirmation Example page and everything went OK until I got to the final part  "Confirm Both Parents in Common Ancestral Couple."

I don't follow how one X match makes it OK to mark both parents of the common ancestral couple as DNA Confirmed, since only one of the pair contributed the X in question. I think all I can do at this point is to put this in the bio of the two children:

• A 108.0 cM match between John Kingman GEDmatch T782948 and his second cousin once removed Kelly Miller GEDmatch A721343 confirm that either the maternal or the paternal relationship can be confirmed by DNA. 

The two children are: Elmer Cleveland Babst and Albert Bruce Babst.

Am I missing something?  

John A. Kingman

Comments

Hello John,

The original question relates to autosomal DNA for 3rd cousins or closer (not X-DNA).

Sincerely, Peter

---

Perhaps it should say that.

Sincerely, John

---

John, I agree with your reasoning. Though in your case, wouldn't an X-chromosome match at least confirm Martha Lanning as Elmer and Albert's mother? She's the only one who contributed X-DNA to them.

---

Of course, you are right. The father can only contribute an X to a daughter. So it would only be the mother of the ancestral couple that could be confirmed in this scenario.

Thanks,

John

Answer 9

Hi,

I have a known second cousin (my maternal side) who has DNA tested, and shows as such, but unfortunately denied this as being the source of the match having disowned their immediate family and claiming to have been swapped at birth... I am no longer in communication with them, but the match has been confirmed subsequently by my first cousin (K.B.) who has since tested, and of course also shows as a second cousin to this person.

Technically I could use the match between myself and the second cousin to DNA confirm back to our great grandparents, but morally speaking it feels a bit wrong to go against their feelings by effectively making the link public even while keeping them anonymous.

Happy to discuss privately with any of the experts on here...