What do you suggest when Y-DNA reveals the shared direct paternal line ancestry is incorrect?

Question

WikiTree shows two males share the same direct paternal line ancestor.  Let’s say they are both 5th cousins to each other on their direct paternal line.  However they only match on 4 out of 25 Y-STR markers.  So one (or both) of their direct paternal lines is incorrect.  

What suggestion(s) do you have on how to indicate that the Y-DNA is showing that some part of the direct paternal line is not accurate?

My thought is additional Y-DNA testing of more recent direct paternal line cousins is necessary to determine in which generation the Non Paternity Event occurred.  But until that happens, should there be a caution/warning sticker on the profiles for each father/son on the direct paternal lines back the shared ancestor? Something else?

Comments

Peter, you haven't explained whether the two males you mention are related to you (or maybe one of them is you) or just someone you came across on WikiTree and happened to notice.  I looked at your profile and you have your own paternal line DNA-confirmed all the way back to your earliest known Roberts ancestor so I'm confused.

On WikiTree, parents can be marked as uncertain, and research notes or "DNA Notes" can be added to profiles.  Someone who observes DNA results inconsistent with the reported family line of another person can contact the PM(s) involved, and recommend more testing as has been discussed below.  But if you don't know where a possible NPE occurred, and it is not your own line you are questioning, I'd be reluctant to take aggressive steps, and certainly not disconnect anyone's parents at this point.

---

Hello Julie, The two males are unrelated to me.  Thanks and sincerely, Peter

Answer 1

I've noticed a similar obvious discrepancy on some profiles and am surprised that the managers do not address it.

My suggestion would be to review the source information and then make a decision.  If the source information is not clear, then that relation should be either disconnected or marked as uncertain.

At a minimum, at least one of the relationships should be marked as uncertain, although it's possible that there is more than one.

Comments

One issue is that a lot of genealogists do not understand how DNA works, plus the impulse to believe what they need to believe to preserve what they think they know. We had an incident here on G2G where men, ostensibly descended from the same male ancestor, had totally different YDNA. Think I vs. R1A. They had convinced themselves that since YDNA mutates, an I could spontaneously mutate into an R1A! Yes! So much more likely than an NPE in their tree! Never underestimate the human ability to see what they want to see.

The really obvious one on Wikitree is the Stewart family. Of course, everyone with the last name Stewart fancies themselves to be descended from the royal Stewarts. On Wikitree, in the DNA connections for the royal Stewart family, a test taker is included who's YDNA is a completely different haplogroup than any of the others. One is reminded of the proverbial ostrich that shoves it's head into the sand, it's rear end wagging in the air.

---

You know, Jessica, it is really depressing (for me) to find someone else shares the same views about Human Nature and what Humans are capable of believing even in the face of stout facts. LOL ​​​​​​​

 to believe what they need to believe to preserve what they think they know. 

---

Susan, just spend a few weeks reading G2G posts about DNA, and you will realize how many people haven't taken a science class since the 1950s. There are people on here who earnestly argue that they didn't inherit any DNA from their maternal grandfather, and that's why their Native American ancestry doesn't appear on their DNA test. Forgetting about the other 22 pairs of autosomes humans have apart from their sex chromosomes!! I guess they think all the DNA we inherit is on our sex chromosomes and mtDNA? The real answer is they're not part Native at all, and that's why it doesn't appear on their DNA test, but they'd rather imagine they somehow failed to inherit any DNA at all from their maternal grandfather! That really makes sense in their heads.

Don't even get into X chromosome inheritance and inbreeding. People on G2G persist in believing that inbreeding somehow degrades the DNA (maybe they're imagining making photocopies of a photocopy, resulting in a faded image?) and they don't get the complexity of things like recessive genes, X-linked disorders, or trisomy.

---

omigosh, Jessica, you're talking about me - the only science classes I've taken since the 50's have been in physics and a variety of engineering disciplines (if you want to stretch to calling that science), none of which have anything to do with DNA.  That doesn't mean that I'm prohibited from learning by reading on my own, though, I hope.  In spite of my lack of competence to figure it all out, I am dedicated to honesty in the facts of who is related to whom.

---

Hey that sounds like most of the Democrats in Washington and the Main stream news . I  know not appropriate here .

In response to Susan Smith comment ..

---

Ah, Terrance, that's possibly -- who knows -- that you may associate the two ideas (for reasons that only you may be able to discern) of Belief and Political Self-preservation. Interesting .... but you may someday discover why the two ideas are associated in your own mind ... Merry Happy Holidays and peace be unto you

---

And MERRY CHRISTMAS to you

---

Actually, Terrance, truth be known, it sounds like 99.99% of politicians of every size, shape, and flavor, not only in Washington, but in every country of the world.

On second thought, it also fits all varieties of extremists … I wonder if there's a relationship between politicians and extremists.  Maybe the connection finders can get to work on that<grin>

---

NOT to be Political, but yes, Belief is essential to the individual support of any Ideology. Belief in 'something' be it the Cause, or its remedy or its sanction or any personal advantage in following that Cause or Ideolgy or ... and etc

And noted that if something is repeated (often enough) it will gather Believers unto it. No matter what it is that is repeated. In Scripture it is said "they have itchy ears, following what they want to hear, that pleases them" (paraphrase) 2Tim4:3

I've found from observation, that applies to any Media pro or con and any worldly field / occupation / educational system and .... etc

And that ain't being political, that's being Philosophical

Answer 2

I have applied DNA genealogy in my quest to find my Fleming patrilineal immigrant to America. I have been in two large Y DNA studies over the past 20 years and studied other Fleming patrilineal lines and their Y DNA in order to sort out families of the same surname who lived in the same region in order to sort them out. Note that you will find that there is never a single Y DNA profile for a given surname. For my surname Fleming, I found at least 4 going back to early America that differed in haplogroup and markers completely, yet lived in the state of Georgia during the same period. So, among Fleming men who immigrated to America, there are several Y DNA genetic lines that make them no more related than a Smith to a Jones. This is for several reasons, but suffice it to say there is no single Y DNA profile for each surname.

Looking at data for hundreds of Fleming men tested, I was able to come to the following conclusions about Y DNA genealogy.

1. DNA doesn't lie. If the results are unexpected, you must seek alternative explanations.

2. If a test subject shares a common male ancestor within 15 or so generations, you will be in the same haplogroup and your markers will differ by no more than 3/67 or 4/111.

An example is this. I found two living male subjects who were direct patrilineal descendants of two brothers of my great x3 grandfather who was born in 1800. Each differed from me by only 1 marker out of 43 but 2/43 with each other. We all 3 had the same haplogroup, of course.

Another two important examples of unexpected findings are these:

1. I tested 2 living male direct patrilineal men who descended from two different sons of a brother (by historical records) of my Fleming great x4 grandfather. They matched each other but did not match me (not even close), but they did match other subjects with the surname of Waggoner. Obviously, we can assume he was likely fathered by a Waggoner male. I checked and there were Waggoner men living close by and interacting with my Fleming family according to documents. Nonetheless he functioned in every way as a part of my Fleming family.

2. I tested 3 living male direct patrilineal men who descended from three different sons of a brother (or cousin) of my great x3 grandfather Fleming. Again, they all matched each other but did not match ANY Fleming in the Fleming Y DNA database. Again, though treated as part of my family and given my surname, he was fathered by a non-Fleming man.

Finally, a number of subjects who have the surname of Fleming who I connect with historically and many that are likely related but not a documented relative show matching Y DNA results with me and other known Fleming men as expected. So, deciding if you are a patrilineal match is not a difficult observation to make.

So, it is clear that unless Y DNA is the same haplogroup and very very close in marker matching to others with your surname who are likely family members, they are not in your direct patrilineal line within 15 or so generations.

If haplogroups and surnames match but markers differ more than 3-4, you do share a common male ancestor in my opinion, but that ancestor likely lived in his home country during the pre-American period.

I hope sharing my years of experience looking a lots of data helps put this important subject into perspective for fellow genealogists. I will also add that I consulted several DNA experts along the way who helped inform my opinion and experience.

Comments

2. If a test subject shares a common male ancestor within 15 or so generations, you will be in the same haplogroup and your markers will differ by no more than 3/67 or 4/111.

That isn't completely accurate John. If you have a recLOH mutation within your Y DNA, which I and one other in my DNA project do, your markers may fall outside of 3/67 or 4/11 well within 15 or so generations. At 12 generations from our common ancestor I match confirmed cousins at 8 and 9 genetic difference in 111 markers and 6 and 7 at 67 markers. I also match at 7/111 and 4/67 with confirmed cousins at 8 generation to our common ancestor. All of us share the same Big Y confirmed terminal SNP (3 member tests). My recLOH cousin matches me at 0/43 markers.

Other than that I agree with the points you make.

---

John,

Thanks for your response. What I was suggesting is a "rule of thumb" for the purpose of context based on studying many lines within the Fleming surname. Obviously, there will be exceptions to such generalizations. My point is that if a patrilineal family member deviates substantially from a close match (ex. different haplogroup or very few marker matches), it is unlikely to be a genetic match that has had significant mutations. It is far more likely to be an ancestor that was fathered through a different patrilineal line. But, as in your case, if you have a solid documented foundation and the markers fall outside of such parameters, but only to a small degree; the obvious conclusion is a true match with more mutations than usual. It should be noted that markers (alleles) mutate at differing rates. All of this context becomes important when you DON'T have a good documented historical connection and you have so-so matches. Understanding how various near and not-so-near matches occur can be helpful in connecting with the "trunk" rather than random "branches" of a family tree. Finally, I often find that people assume that men with the same surname are automatically expected to match in Y DNA. The truth is that men with the same surname often have completely different Y DNA for varying reasons.

Answer 3

I would suggest you run their results through the Nevgen predictor. If the predictor predicts they are in the same haplogroup, you  should remember that STRs mutates randomly and have had 12 chances to mutate. It would be rare to have a GD 4 at 25 markers with 5th cousins but not impossible. 

If they are in the same haplogroup, then BigY for one and single SNP testing of the other for the terminal SNP.

If they are not in the same haplogroup or the other cousin is not derived for the terminal SNP, then BigY of both and single testing of other cousins for the now 2 terminal SNPs

Comments

They only match 4 out of 25 markers, so it is a genetic distance (GD) of 21.

---

Oops, sorry. Then the second option

Answer 4

It seems to me your question concerns how to note on Wikitree profiles when there is an obvious NPE in someone's lineage as proven by Y-DNA.  If there's a Y-DNA surname project for the particular line, then you'll know which one has a legitimate connection and which one doesn't.

I do think it's a good idea to have some sort of sticker on the profiles up the patrilineal line.  Additionally, the paternal relationship for each generation up to the purported shared common ancestor should be marked as uncertain.  I'd suggest you come up with a proposal for the wording (and image) for the sticker.

This sticker would then be placed on the father of the tester whose Y-DNA isn't matching those in a Y-DNA surname project as well as up the patrilineal line to the son of the common ancestor.  In the case of two individuals with no surname project, the sticker would be placed on the profile of the father (and patrilineal ancestors up to the common ancestor) of both testers.

Comments

This might run afoul of the rule that DNA test takers must manage their own profiles in order for the test information to be on WikiTree Darlene. If a profile is marked as DNA confirmed without the test information I believe it trips an error flag. Any sticker leading from an ancestor down to a descendant who does not manage their own profile, and thus does not show the DNA test information that descendant took, is unverifiable.

As long as all of the test takers involved manage their own WikiTree profiles there shouldn't be a problem, but that might not always be the case.

---

So you're proposing to place a sticker on the profile of a WikiTree member's father stating he's not the birth father?

---

Peter would have to respond, since he asked the question, but my interpretation of his question and situation is when two people on Wikitree have posted their Y-DNA test result with haplogroup and have connected their line back to a common ancestor.  Since the two males have a different haplogroup, then they do not share a common patrilineal ancestor.  If one (or both) of them is part of a Y-DNA surname study for the surname, then you know if they do or do not match the haplogroup for that common ancestor.  If there isn't a Y-DNA surname study/group, but both men are claiming a common patrilineal ancestor, we know one of them doesn't in fact descend from that ancestor.  We just don't know which one it is.

No one is talking about managing someone else's profile or putting anything on a test taker's profile, so I'm not suggesting anything that runs 'afoul' of any rule...  The sticker Peter mentioned would be on the descendants of the common patrilineal ancestor, up to but not including the test taker.

Julie, I'm not proposing anything.  Peter started a G2G discussion and asked for comments.  I commented that I thought a sticker to make others aware of a 'non-match' as to Y-DNA would help prevent misunderstandings.  There are a lot of people that get DNA tests that don't have any clue what the test results actually are telling them.  Certainly the sticker wouldn't claim anyone's father wasn't their father.  I don't know what phrasing Peter had in mind, but I would guess it would be along the line of mentioning that the Y-DNA test was inconsistent or created uncertainty as it related to 'such and such' (with 'such and such' being the common ancestor).  As stated, it's not known in which generation the NPE occurred.  All that's known is that the Y-DNA has confirmed there is one in someone's line.

---

Thank you, Darlene.  I have been thinking about this question ever since my last post.  Please tell me if I'm wrong, but here's the way I see it:

Take as an example my father, W.D. Kelts.  His line is Peter Kelts > John Kelts > George Kelts I > George Kelts II > my father.  If my father, who died long ago and was never tested, was not a Y-DNA match to another descendant of Peter, that in no way invalidates the whole line.  Maybe my father was not the son of George II, but George II could still very well be the son of George I, etc.  

As far as I know, WikiTree does not make a practice of marking every parent as uncertain just because the relationship has not been confirmed by DNA.  Thus it seems to me to be an extreme reaction to say (and I know you're not the one who said it) that each parent all the way up the line should be marked as uncertain.

I absolutely agree that many people don't understand DNA.  But if one person at the end of a line ignores DNA results that brings his parentage into question, I don't see how that presents much danger to the overall shared tree.  In normal cases (excluding people who have their hearts set on being descended from royalty, etc.), I would not expect a whole group of a particular person's descendants to ignore DNA and insist on disproven connections, so I doubt the error could go very far back, at least not in the long run.

In this particular case it looks to me like a couple of people have called the original conclusion into question, although not having studied up on Y-DNA lately, I don't have the expertise to evaluate their statements.

---

I would not expect a whole group of a particular person's descendants to ignore DNA and insist on disproven connections . . . .

I have that very situation in my mother's paternal lineage. In that line a traditional paternal line has been disproven by DNA and proven to be connected to another line, though by SNP results rather than STR (both paternal lines are the same haplogroup R1b). A number of descendants have fought, and some continue to fight, the change from the traditional lineage to the DNA proven lineage. 

The SNP results are definitive while STR results are predictive, and both show the error in the traditional lineage. There are a large number of descendants of both lineages who have tested their Y DNA, including some from each lineage who have taken the Big Y test. The test results are very conclusive, so those who continue to resist changing to the DNA proven line now insist that the traditional lineage could be correct through adoption or fostering and such.

My point being that you cannot predict how people will react when presented with DNA data that contradicts their long held tree beliefs. If, or when, there is a conflict between what the DNA reveals and beliefs that have long been held, tread carefully and remember the advice not to Wiki when angry.

---

Interesting, John.  And now that I think of it, I do have examples in my own lines of people ignoring evidence and refusing to change disproven connections and other information (in my cases, it is other evidence and not DNA, and it is not large groups of descendants).  

One good thing about WikiTree is that there are mechanisms--mentors and mediation--that can be resorted to in such cases so that ultimately we can achieve a shared tree that is consistent with the evidence.

Answer 5

I ran across this problem a while ago. When I saw that three men were all descended from the same man, but one was I, one R1b and one R1a, I posted a note that there was something wrong. I just went and looked and the R1a has disappeared, but the I and R1b remain. And I suspect the R1a disappeared only because of the more stringent DNA posting requirements that were implemented this year. So, sadly, I do not have an answer to your question, but agree that it points to a problem.