Should the Y DNA Confirmation template be changed now that the TiP changed at FTDNA?

+4 votes
396 views

The Y-DNA  "Confirmed with DNA" source citation includes the Genetic Distance and the probability the two matches share a common ancestor within a certain number of generations. FTDNA updated the TiP and appears to only now give you the GD but not the probability of a common ancestor within a certain number of generations. Perhaps that information is still available at FTDNA, but I haven't found it.  Here's the current WikiTree example for the Y-DNA Source Citation to support the father's relationship "Confirmed by DNA" button. Should this be updated to use information one can find at FTDNA?

* Paternal relationship is confirmed through Y-chromosome DNA testing. [[Smith-44676|William Waugh Smith]], FTDNA kit #95454, and his 6th cousin once removed, [[Smith-81508|Len Smith]], FTDNA kit #216226, match at a Genetic Distance of 2 on 67 markers thereby confirming their direct paternal lines back to their MRCA [[Smith-15895|Thomas Smith]]. FTDNA indicates that the probability the two share a common ancestor within the last 7 generations is 91.73% and within the last 8 generations is 94.73%.

in Policy and Style by Terri Stern G2G6 (7.1k points)
Can someone from the DNA team answer my question? The format for the Y DNA Confirmation example is no longer supported by the change FTDNA made to the TiP. Don't we need a new example format?

Just a comment to note that Terri's original Question is almost a year old now, but the Help:Y-Chromosome_DNA_Confirmation page hasn't been updated since July 2022.

Since the instructions about use of the FTDNA TiP (Time Predictor) report and the example "Confirmed with DNA" citation shown haven't been correct since FTDNA changed the TiP format in February 2023, I do think it would be a good idea to revisit that Help page.

Hello Edison (and others),

Would you please help me find two males with the same or similar surname who are both:

* In FTDNA’s “GFR + WikiTree” project (So I may see their TiP report)

* and have (preferably) tested 67 or more Y-STRs

* and have a known patrilineal relationship in WikiTree (preferably) greater than 5 generations

* and are willing to have their names and relationship used as an example for the help page.

Thanks and sincerely,

I'm afraid I can't help with that, Peter; there are a lot of conditional criteria in there. smiley

But the real point is that the TiP report, since February 2023, no longer works the way it's described on the WT Help page. It's now a static chart--it doesn't change from match to match--and there are no parameters that can be set (i.e., number of generations for which to display a value...in fact, there are no longer generational estimates displayed at all, only a 95% CI range and a median year) and no percentages are shown for probability weighting.

All the TiP report does now is display the same chart for each instance, but with the appropriate timeframe highlighted based on the number of STRs tested and the calculated genetic distance.

Here are two "live" examples I just pulled, both using the 67-marker level of testing (click the images for larger versions). The first is for two men with the same surname who are a genetic distance of 1:

The second example shows men with different surnames and a genetic distance of 7:

I don't know that a different "Confirmed with DNA" citation example using different test kits would help...other than wanting to find one where both kits are in the FTDNA WikiTree Project and who both have WikiTree profiles and who consent to being named for the comparison.

Fundamentally, the issue is that the entire structure of the TiP report changed a year ago and the instructions and example of the Help page for yDNA "Confirmed with DNA" status simply can't be applied as described.

BTW, as an aside, I've always wondered why the WikiTree FTDNA Project doesn't sort kits into genetically similar groups, as do most yDNA projects. Everyone is lumped together in just a handful of headers, essentially just the basal haplogroups. It's sort of impractical for anyone to decipher who matches whom without capturing the entire results chart, plugging it into a spreadsheet application, and trying to sort everything oneself. Why was that approach taken?

Edited: Because I actually went back and reread one of my abysmally worded sentences. I rite gud... frown

Does the highlighted timeframe provide supporting evidence for the known patrilineal relationship?

Peter, I believe--personal opinion only--that some of the TiP report's timeframes do not offer adequate supporting evidence to be practical for genealogy...but I've always been on the more conservative side of the spectrum for interpretation of the pertinence of DNA evidence.

The older TiP report came under almost constant critique because it was so generalized and a comparison at or above a 90% generational range could encompass a very broad range. Frankly, the new TiP report has probably dumbed that down even further, but then again using only Y-STRs has a limited usefulness for relationship prediction anyway.

My opinion is that the confidence interval ranges on the new TiP report can't do a great deal to support a hypothetical MRCA even with decent genealogical information to support it. The absolute best scenario is a 111-marker match at a GD of zero, but that still only affords a range of 200 years to the MRCA.

Ignoring the Y-12 and Y-25 tests, the winner for the not-much-help award is a 67-STR "match" at a GD of 7: the CI spans a whopping 1,200 years. Kinda like saying, "Yes; I am a descendant of Charlemagne!" (and Charlemagne was born 350 years after the start of that CI range on the TiP report).

What I wish FTDNA would do is tighten their threshold for reporting STR-only matches. IMHO, 37 markers should go out to only a GD of 3; 67 markers to GD 5; and 111 to GD 8. Or possibly report them, but not as an indicator of a positive match within the genealogical timeframe. A 37-marker test that reports a "match" that could have a common ancestor born as long ago as 650 CE is sort of pointless.

It won't be a popular opinion, but I believe that if WikiTree limited use of Y-STR evidence to those genetic distances for a "Confirmed with DNA" status (37, GD 3; 67, GD 5; 111, GD 8) it would at least keep the CI ranges at levels that might have a reasonable possibility of supporting paper-trail evidence within the genealogical timeframe. A window as deep as 400 CE or 650 CE is pretty much pointless for pedigree work.

An additional aside: If the Help page for yDNA "Confirmed with DNA" is going to be revisited, might not this also be a great opportunity to finally added some specific Big Y confirmation guidelines and examples? That's really the gold standard right now for yDNA matching and generational prediction. The Big Y first came out in 2014 and the Big Y-700 in 2019. It would be great to see WikiTree bring the yDNA Help page up to date for that technology.

Hello Edison,  Imagine your second cousin once removed Clifford Nolan Williams https://www.wikitree.com/wiki/Williams-77118 had a son named John who took a Y-DNA67 test and matched you on 66 out of 67 markers.  The TiP report would highlight 1600 - 1900 with the estimated TMRCA of 1800.   The most recent direct paternal line ancestor between you and John is Elijah Williams https://www.wikitree.com/wiki/Williams-51876 born 1790.

Your 66/67 match with Clifford's son is supporting evidence that each father / son relationship between John and Elijah, and you and Elijah, is accurate.

I'm afraid I've lost the logic of the discussion, Peter. I noted above that--just my opinion--usable STR-only evidence would probably be meaningful out to GD 3 for 37 markers, to GD 5 for 67 markers, and to GD 8 for 111 markers.

I'm not certain why you'd pose the hypothetical scenario of a GD 1 at 67 markers with a known paper-trail 2C1R.

On that specific Williams line, we have a 22-year-old yDNA subproject that has over 35 participants, 24 of them with Big Y tests. I am, in fact, in a position to provide supporting evidence to a fairly specific timeframe for just about any Williams male whom FTDNA reports being as distant as even a GD 4 at 37 markers, provided they have a solid pedigree tree back at least a few generations. We happen to have two STR mutations within the 25-panel results and one within 37 markers that are unusual in the M-269 subclade and, combined with the depth of Big Y testing we have, the generational coalescence points are clear to within a single generation. So I'm not a good hypothetical.

If, however, we're talking about a general approach to be used by even the casual genealogist who does not have the kind of DNA data I do, I'll stand by my statement that, for example, a GD of 7 at 67 STR markers in and of itself will not provide a useful form of positive supporting evidence. That puts the median TMRCA at before the time of consistent surname adoptions in western Europe.

Again, though, the subject of Terri's original Question is that FTDNA changed the TiP report a year ago, but WikiTree's yDNA "Confirmed with DNA" Help page has not been updated since 18 July 2022. Regardless of what I or anyone else thinks the parameters should be to establish minimum supporting evidence, the core issue is that the Help page is providing instructions and examples that no one has been able to follow for a year now.

"...usable STR-only evidence would probably be meaningful out to GD 3 for 37 markers, to GD 5 for 67 markers, and to GD 8 for 111 markers."  Yes and there are instances of 4 for 37 and 7 for 67 and 11 for 111 when the patrilineal cousins are more distantly related.  In WikiTree we are (should be) dealing with documented father / son relationships.  Is the genetic distance between patrilineal cousins in agreement with their known relationship? If so, then that is supporting evidence that their direct paternal lines back to their shared patriarch are accurate.

Thanks, Peter. But I feel like we're discussing--or attempting to discuss--two different things. The relative merits of using Y-STR data--in the absence of additional genetic information--as a form of positive evidence for genealogical, pedigree hypotheses is a complex subject, and one we could debate separately. But it isn't the subject of Terri's Question from 10 months ago.

Terri's Question, and the topic at hand, concerns the fact that FTDNA revamped their TiP report in February 2023. The instructions at Help:Y-Chromosome_DNA_Confirmation have been incorrect since that date, and it is impossible for WikiTreers to explicitly follow the instructions as written.

Because WT Help pages can be edited only by a select few individuals, it seems it would be the DNA Project who should call their attention to matters like this and recommend changes to the text in order to keep the instructions current and applicable.

Corollary to this (I can't tell since I don't use it) I believe Greg Clarke's handy "DNA Confirmation Citation Maker" app may need to be modified, as well. But I think Greg works off the specific directions and parameters he derives from the WikiTree Help pages for "Confirmed with DNA," so it seems the fundamental gap remains those incorrect yDNA instructions.

To update the template I need an example I can see.  If the matching males are both in FTDNA’s GFR + WikiTree project then I can see their TiP report. Thanks.

Peter, given that the GFR & WikiTree DNA Project is not grouped by genetic relationships but only by basal haplogroups, I think it would be incumbent upon the administrators of that project to locate an example match.

I have only a public view of the DNA project's results, but I'll go through a sorting and matching exercise for you to see if we can arrive at some examples. Of course, you'll need to use your admin GAP login to identify the WT IDs. But let's see...

  • Kits B8311 and N71131; at 111 markers, GD 1
  • Kits 971432 and B811455; at 111 markers, GD 1
  • Kits 407095 and 17236; at 67 markers, GD 3
  • Kits 303002 and 528892; at 67 markers, GD 1
  • Kits 156283 and 117039; at 37 markers, GD 1
  • Kits 86879 and N2642; at 67 markers, GD 1
FWIW, at a first-glance sorting and comparison, only 83 kits among the 1,289 project members can be reliably assembled into matching genetic groups. So I don't believe you have a huge number of active WikiTreers from that project that could be potential examples for a revised citation.

It would only be to 37 markers and he does not have his kit number displayed at the GFR & WikiTree DNA Project, but couldn't you use your 2C1R, Thomas, and yourself as the example?

1 Answer

+4 votes
 
Best answer

I agree with Terri that this needs attention. Family Tree DNA introduced its revised TiP (Time Predictor) tool two months ago, on February 16, as explained and described here: https://blog.familytreedna.com/ftdnatip-report/.

The current instructions read: "Does the Family Tree DNA TiP report predict a similar number of generations between the test takers as traditional genealogy? If so, you can mark the paternal relationships as Confirmed with DNA."

Even for STR-only tests, data from Big Y results have been extrapolated to improve the TMCRA estimations. One significant advantage, should WikiTree choose to use it, is that now we have years-before-present and calendar date ranges for 99%, 95%, and 68% CI (confidence interval) ranges.

Working only with STR data can lack precision, especially below the 111-marker panel. I would suggest the "Confirmed with DNA" guideline be changed to the narrower CI range of 68%, or certainly no more than 95%. A CI is simply a computed probability that all such events will fall within the stated parameters 68% (or 95%) of the time.

It's quite likely that the wording in the current instructions of "a similar number of generations" may have resulted in DNA confirmations that will now fall outside of a 68% or even 95% CI. We should probably include some guidance on how members can check their previous yDNA "Confirmed with DNA" statuses, and what to do if the new FTDNA TiP data no longer support the previous assumptions.

by Edison Williams G2G6 Pilot (504k points)
selected by Lucas Van de Berg
I just went through this process today and it's extremely confusing. They no longer show the other person's test ID either.  So I wrote what I could and was wondering if you could check it and offer any suggestions?  I'm familiar with atDNA confirmation and I don't think this Y confirmation source statement really shows confirmation.  But it's the best I can do with what FTDNA provides now. You can see what I wrote on https://www.wikitree.com/wiki/Burchett-479.  Thank you!

Hi, Diane. While, at a glance, I don't think there's anything wrong with the statement you added to Burchett-479, what I definitely am not is any sort of authority on WikiTree "Confirmed with DNA" guidelines.

I think John Kingman or someone else from the WikiTree DNA Project who deals with that every day will be by here in the not-too-distant future and can give you an official opinion. laugh

Related questions

+7 votes
1 answer
+6 votes
0 answers
86 views asked Nov 19, 2024 in The Tree House by Russ Carter G2G6 Mach 3 (35.5k points)
+15 votes
2 answers
2.3k views asked Jan 30, 2019 in The Tree House by James Stratman G2G6 Pilot (108k points)
+5 votes
2 answers
311 views asked Nov 3, 2024 in Genealogy Help by NG Hill G2G6 Pilot (126k points)
+8 votes
2 answers
+8 votes
1 answer
+5 votes
1 answer
+2 votes
1 answer

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

disclaimer - terms - copyright

...