Are we further confirming a parent/child relationship or proving a match?

+11 votes
170 views

I believe the statements and terminology chosen in the help can be improved. The use of DNA Evidence in Wikitree seems different than how it is used outside Wikitree.  I am basing what is used in Wikitree and outside of Wikitree on my experience so others may have a different experience or understanding.

DNA is treated elsewhere as additional corroborative evidence with significant weight. Wikitree seems to treat DNA more along the lines of independent Verification. The reason I and many other purchased additional kits for known relatives was to aid in finding connections of unknown relatives. We proceeded under the premise that a match was inevitable. For most, not me, this proved to be true.

Confirmation of something is a belief that something is true and will remain true. It is not proof. Confirmation of a Parent/Child relationship also has a confidence level associated with it.  IMO, the Wikitree use of these levels has been the source of confusion. Generally speaking, there are many degrees of confidence any one individual may describe of a parent/child relationship in a tree.  But when we acting as representatives, as on Wikitree, it is good to establish a limited set of defined choices. 

Wikitree has three levels of Confidence associated with a parent/child relationship.

  1. Uncertain
  2. Confident
  3. Confirmed with DNA

For the purpose of this example, I will limit the choices outside of wikitree to only two comparable levels 

  1. Uncertain
  2. Certain

And some indicator that DNA Evidence exists.

The main difference is that Wikitree guidelines do not provide a choice to indicate (1) (a)uncertain and (b) corroborative DNA Evidence exists, or Confident and corroborative DNA exists but the DNA Evidence does not meet requirements of “Confirmed with DNA”.

A typical scenario, outside Wikitree, involves two matches, Jack and Jill, who have been able to identify their Most Recent Common Ancestor(s) via parent/child relationships that are certain. A presumption is made that corroborative DNA evidence exists up to and including their Most Recent Common Ancestor(s).

A third person, Bill, matches both of them and does not have a completed tree but does share a triangulated segment with the two. In this example, let us add that all 3 participants are predicted to be 4th cousins.  The evidence strongly suggests that Bill also shares those same Most Recent Common Ancestors.

The difference between Wikitree and NonWikitree users is the requirement of a 3rd tree and triangulation. It’s worth repeating, Confirmation of something is a belief that something is true and will remain true.  

In this example, if I were to “Confirm with DNA” the parent up to and including the Most Recent Common Ancestors of Jack and Jill. I am stating that I am still confident the parent/child relationship and that no 3rd DNA test will refute this. 

Why am I so confident? The reason is that NO Third autosomal test can refute it.  The confidence level remains virtually the same in all cases.

When is a 3rd tree and triangulation required?

A third tree and triangulation is required for determining the parent-child relationship of a segment, not a person. In the example with Jack and Jill, we can trace these segments up to but not including the Most Recent Common Ancestors.  Let’s change the relationship of Bill from another 4th cousin to a 5th cousin.  The issue, in this case, is that Bill didn’t care where the segments came from, as long they came from the same MRCAs of Jack and Jill.  The prediction enlarges the possibilities to the Parent or Parents of Jack or the Parent or Parents of Jill. 

The third tree with another cousin of a more distant relationship would help to identify which side of Jack and Jill’s MRCA the relationship connects.  A triangulated segment helps to identify the parent/child relationship of a particular segment.

In summary, an additional DNA test and documented tree cannot refute the results of the 1st match then why require it?

Edit: I edited the question to reflect topic

in Policy and Style by Ken Sargent G2G6 Mach 5 (57.0k points)
edited by Ken Sargent

2 Answers

+7 votes
Hi Ken, This all went way over my head.  Can you simplify your question?  We are only dealing with one family tree here, so the twigs either meet on a branch at a common ancestor, or they don't, yet.  

I think your question seems to be why is triangulation necessary to use the Confirmed with DNA tag. I think that triangulation is the only way to verify the traditional research and confirm the initial DNA connection.  Otherwise, each parent-child relationship is either Uncertain or Certain.  I think of the Confirmed With DNA as confirmation of the initial DNA connection by one or more matching DNA tests.

And yes, WikiTree does approach this differently than other websites because this is a one world tree.  We are confirming individual parent-child relationships that have been matched with more than one DNA test by way of a common DNA ancestor. If the tradional research is incomplete, the parent-child relationships can only be confirmed as far back as the traditional research goes. . . I think. . . but I may be confused.  :-/
by Kitty Smith G2G6 Pilot (557k points)
The parent-child relationships can only be confirmed as far back as the most recent common ancestor. See: http://www.wikitree.com/wiki/DNA_Confirmation#Y-Chromosome_Confirmation

If the traditional ancestry is incomplete for Y-DNA tested direct paternal lines you may use a Y-STR category (e.g. http://www.wikitree.com/wiki/Category:Roberts_Y-STR_Group_1).  For mtDNA tested direct maternal lines, you may use a mitotype category (e.g. http://www.wikitree.com/wiki/Category:H1a3a1_mtDNA_Group_1)

Using auDNA and cousin matching out to 3rd cousins you should use a quantitative approach (total amount of shared auDNA is in the expected range for the relationship).  Beyond 3rd cousins triangulation is necessary.  See http://www.wikitree.com/wiki/DNA_Confirmation

Sincerely,

I was a bit confusing so let me try this....

"The "Confirmed with DNA" indicator means that traditional genealogy has been further confirmed with DNA testing" 

see Does "Confirmed with DNA" mean a relationship is proven?

I believe the current requirements/restrictions documented in determining if a parent/child relationship has been further confirmed with DNA Testing has a different undocumented objective.

IMO, a single DNA Match between two known relatives does further Confirm with DNA some profiles that connect these two, 

IMO, a requirement that only 3 confident trees with 3 DNA tests that the 3 tests must have at least one single overlapping segment go well beyond the stated objective.

IMO, the ability to identify profiles that further Confirm with DNA is important and it has the added advantage of fitting within the existing traditional genealogy concerning evidence.

If the DNA evidence further confirms with DNA evidence, then the option should be selected, if it does not then it should not.

+4 votes
To me the reason is simple, as pointed out by Peter Roberts the shared DNA percentage can only be used for closer relationships. What I think is easy to mix up is matching and being related by DNA within the recent genealogical time frame. In other words, IBD vs IBC (or IBS).

Unless you prove through triangulation the match between the supposed to be 4th cousins on paper is not proven IMO. People tend to forget that a 2.5% of average (meaning every 40th child birth) is a NPE.

That's why we do DNA tests, to prove that our paper trail is correct and identify where it's not correct.

I do agree to most of your points though and I now understand much clearer where you're coming from. Still, a 7cM and 700 SNP segment has a chance to be IBC. Having a paper trail would also make me lean very clearly towards IBD but still to be almost 100% (as on these levels the ancestral segments might be lost or there is no third person right now but maybe in the future) sure there is the need to do a triangulation.

Unless there is another category just below "confirmed with DNA" which articulates that DNA evidence is indicating a relationship and is supporting the paper trail.
by Andreas West G2G6 Mach 6 (60.6k points)

Andreas, Thank you for your response and I will address your points one by one.

1st, I am not sure why % shared is used, at least alone.  It is considered the least accurate. This is why DNA Services also use the # of segments, and algorithms to address anomalies found using their extensive database. They are the experts in this area and believe considered and authoritative source.

2nd. If the match is between 2 tests from two different DNA Services and Gedmatch is used, Gedmatch does not report % shared.

I am glad you mentioned NPE because this will allow me to identify at least one area of agreement. Let us use the example of a known and well-documented 4th Cousin whose predicted relationship is consistent with all 3 DNA Services.  I realize that this may not be something you would accept, so let us say a valid match consistent with someone who is a 4h cousin.  

We are not attempting to prove these two people are 4th cousins which I believe involves a one to one connection. We are actually deciding if the "traditional genealogy has been further confirmed with DNA testing" on up to 10 Profiles.  8 of those profiles asked the question about 1 parent. These are the profiles up to but not including the Most Recent Common Ancestor(s). 

2 of the profiles asks the question about 2 parents(normally). These are the 2 Children of the Most Recent Common Ancestor(s).

With a valid match, we can presume to know the parental source of these segments (in normal cases) for each of the DNA Testers profiles up to BUT NOT including the Most Recent Common Ancestor(s).  For each of these 8 profiles, we ask the question about further DNA Evidence.

IMO opinion, the answer is yes and so we should check the box.  In other opinions, the answer is no and a 3rd test with a confident tree connection and a triangulated segment must exist to prove the parent/child relationships exist for these 8 profiles.

There are some opinions that these are uncertain and that the a positive third test as described effects the conclusion so dramatically, we can then confirm the parent of each of these 8. My understanding is that a 3rd test does not affect them in any way.  

Can someone please explain to me or point me to some paper as to how a third test accomplishes this?

We are in agreement that we can't confirm either Parent of the 2 MRCA's children.  This is because there may have been a NonPaternal Event. We can not tell if these Children shared 1 parent or two.  If they shared only one parent, we don't know which one.

The only way to know if a parent can be confirmed with DNA is via Triangulation.  NOT every match that triangulates can give us this answer but for those that do, the result affects ONLY these 2 profiles on one parent each. It has absolutely no effect on the 8 descendants. 

if the result does effect these 8, please set me straight by explaining how.

I have said something like this in the past but let me try this another way.

The current guidelines are based on the current guidelines used help resolve brick walls  Genetic genealogists map chromosomes using the shared segments of known matches. A third match, without a known connection, uses triangulation to find a connection.

Using match information, genetic genealogists can associate each segment of DNA up to a Common Ancestor if only one exists, in cases of 1/2 cousins, or up to but not including either Common Ancestors, in cases of full cousins.

A confident relationship includes a belief that no other documentation will change this parent/child relationship, but DNA might reveal the documentation is wrong by finding a NonParental Event.  

This is where Wikitree and I differ on the value of any 3rd DNA test For someone mapping the genomes, Each additional match adds more segments that can be mapped but if the segment is triangulated, it adds absolutely no value at all for those profiles up to by not including the Common Ancestors.

No third autosomal DNA test can reduce the confidence, it can only increase it. For the Genetic Genealogist, it would seem the least desirable match is one that triangulates between known DNA relatives at the same or closer levels because it adds nothing new.

Triangulation is required in the context of proving relationships that exist between two DNA Testers.  That is a different subject. I would like to focus on the parent/child relationship of those profiles up to but not including the Common Ancestors.

For these Profiles, there is virtually NO chance that a NonParental event occurred on that line and NO Chance that a 3rd test could indicate differently.

My last attempt for a while...:)

I believe it's worth the time an effort to understand that you can virtually guarantee that a Non-Parental Event is NOT going to occur (with the exception a close relative), for every parent/child relationship on every Profile up to but not including the common Ancestor(s).  

This is true in cases of a well-documented connection between themselves and another relative with a DNA Services prediction consistent with these trees.

This means that if a Genealogist with basic level skills can confirm, within the stated limitations, 100% of their matches who meet this documentation requirement and who subscribe to all 3 services.

Related questions

+5 votes
1 answer
131 views asked Jan 6, 2018 in Genealogy Help by Pace Bonner G2G1 (1.1k points)
+3 votes
3 answers
145 views asked Jun 23, 2019 in Genealogy Help by Susan Smith G2G6 Pilot (492k points)
+5 votes
1 answer
+4 votes
5 answers
252 views asked Nov 26, 2019 in Genealogy Help by Katherine Cline-Bowman G2G5 (5.4k points)
+1 vote
1 answer
126 views asked May 29, 2019 in Genealogy Help by Amanda Tomas G2G2 (2.7k points)
+5 votes
1 answer
+4 votes
0 answers
128 views asked Jul 2, 2018 in Policy and Style by Kathleen Cobcroft G2G6 Mach 7 (71.8k points)
+6 votes
1 answer
135 views asked Mar 26, 2018 in Genealogy Help by Chris Ferraiolo G2G6 Pilot (452k points)
+9 votes
10 answers

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

disclaimer - terms - copyright

...