Let's All Start Using Terminal SNP Labels Instead of Y Haplogroup Subclade Names

[razyn]

Have you all seen the following notice on your myFTDNA pages Haplotree page?

Therefore, in the very near term, Family Tree DNA will discontinue showing the current “longhand” on the tree and we will focus all of our discussions around your terminal defining SNP.

This changes no science - it just provides an easier and less confusing way for us all to communicate.

Bennett Greenspan, Family Tree DNA
Dr. Michael Hammer, University of Arizona

I am more OK with this if they let us declare what SNP that is, even if it's based only on FTDNA testing.  Because, as I tried to point out previously on this thread, their version of my terminal SNP is still R-P312, my "Deep clade" result -- even though I have tested six beyond that, at the same lab, and am positive for four of them.  The seriously laggard data in the "My Haplotree" page continues to assert that R-P312 is my terminal SNP, and that there's nothing else to test.

Normal people, by which I mean those who don't spend hours daily following these forums, should be expected to believe that nonsense.  And maybe that will drive normal people to the Geno2 test, I don't know; but if so, is that supposed to be a good thing?

[rms2]

Well, they won't go with the real terminal SNP until that SNP is accepted by the YCC and added to its tree. That's the long and short of it.

[rms2]

I think using "R" alone in short hand designation is too broad . . .

It's not too broad when followed by the terminal SNP. The letter designator takes you to the branch, the terminal SNP directs you down the branch to the twig you're on. It's pretty straightforward. Anyone who cares to look can see you're not R1a.

R is too broad, but R1b is not? Why not R1b1a2-DF23, or R1b1a2a1a1b4-DF23? Then you not only won't be in any danger of someone mistaking you for an R1a or an R2, you won't be in any danger of being mistaken for a xM269 or a U106+.

[eochaidh]

I think using "R" alone in short hand designation is too broad . . .

It's not too broad when followed by the terminal SNP. The letter designator takes you to the branch, the terminal SNP directs you down the branch to the twig you're on. It's pretty straightforward. Anyone who cares to look can see you're not R1a.

R is too broad, but R1b is not? Why not R1b1a2-DF23, or R1b1a2a1a1b4-DF23? Then you not only won't be in any danger of someone mistaking you for an R1a or an R2, you won't be in any danger of being mistaken for a xM269 or a U106+.

Why make anyone look it up? R1b DF23 is pretty easy to write and read. Your signature says R-DF41 (L21>DF13>DF41), which is quite long. I have the feeling it may change soon, though  :)

[rms2]

I think using "R" alone in short hand designation is too broad . . .

It's not too broad when followed by the terminal SNP. The letter designator takes you to the branch, the terminal SNP directs you down the branch to the twig you're on. It's pretty straightforward. Anyone who cares to look can see you're not R1a.

R is too broad, but R1b is not? Why not R1b1a2-DF23, or R1b1a2a1a1b4-DF23? Then you not only won't be in any danger of someone mistaking you for an R1a or an R2, you won't be in any danger of being mistaken for a xM269 or a U106+.

Why make anyone look it up? R1b DF23 is pretty easy to write and read. Your signature says R-DF41 (L21>DF13>DF41), which is quite long. I have the feeling it may change soon, though  :)

They'd still have to look it up either way, if they care enough to know whether it is downstream of P297 or L11 or P312 or U106, etc.

[Mikewww]

R is too broad, but R1b is not? Why not R1b1a2-DF23, or R1b1a2a1a1b4-DF23? Then you not only won't be in any danger of someone mistaking you for an R1a or an R2, you won't be in any danger of being mistaken for a xM269 or a U106+.

Its no big deal but the positive on R1b is that is a well recognized label (in studies, etc.) and the difference between R1a, R2 and R1b could just as well be  X, Y and Z.  In other words, it's only by chance of occurrences 20,000 years ago and chance of the original discovery sequence and lettering scheme that this is R.  On the other hand, at least in Europe, R1b is effectively L23 or even L11.

Anybody can look up anything, but we should try to make our discussions easy for the new folks and casually interested.

[rms2]

I don't think "R1b" gets you much more than "R". R gets you to the branch. The terminal SNP gets you the rest of the way.

The terminal SNP is the thing, and it seems to me important to use the shorthand in the correct way, rather than monkeying with it to suit this or that person's personal preference. We already have too many different ways of naming things in genetic genealogy.

[seferhabahir]

I don't think "R1b" gets you much more than "R". R gets you to the branch. The terminal SNP gets you the rest of the way.

The terminal SNP is the thing, and it seems to me important to use the shorthand in the correct way, rather than monkeying with it to suit this or that person's personal preference. We already have too many different ways of naming things in genetic genealogy.

So what is the terminal SNP going to be for me and for others that are downstream from DF13 but "under investigation" like L526, L580, and L583. Will FTDNA be calling me R-DF13 or R-L583?

[rms2]

I don't think "R1b" gets you much more than "R". R gets you to the branch. The terminal SNP gets you the rest of the way.

The terminal SNP is the thing, and it seems to me important to use the shorthand in the correct way, rather than monkeying with it to suit this or that person's personal preference. We already have too many different ways of naming things in genetic genealogy.

So what is the terminal SNP going to be for me and for others that are downstream from DF13 but "under investigation" like L526, L580, and L583. Will FTDNA be calling me R-DF13 or R-L583?

That's a good question. My guess is R-DF13.

[Mikewww]

I don't think "R1b" gets you much more than "R". R gets you to the branch. The terminal SNP gets you the rest of the way.

The terminal SNP is the thing, and it seems to me important to use the shorthand in the correct way, rather than monkeying with it to suit this or that person's personal preference. We already have too many different ways of naming things in genetic genealogy.

So what is the terminal SNP going to be for me and for others that are downstream from DF13 but "under investigation" like L526, L580, and L583. Will FTDNA be calling me R-DF13 or R-L583?

That's a good question. My guess is R-DF13.

Unfortunately, this will probably be only as good as FTDNA's ability to keep their programming current.  So, really there is no change there unless they've rewritten some of their stuff and their support processes. Right now, it takes over a year to update their haplotree so a year from now DF13 could seem rather ancient.

We do need to pay tribute to the volunteers at ISOGG who keep their tree current. No programming is required to support applications, such as what a testing company needs, but at least ISOGG keeps the charts current.

[gtc]

I don't think "R1b" gets you much more than "R". R gets you to the branch. The terminal SNP gets you the rest of the way.

Sure, if you have a tree diagram at hand to look-up unfamiliar SNPs.

Frankly I'd be happy if R1b were to be renamed something in the range U through Z.

[Heber]

I don't think "R1b" gets you much more than "R". R gets you to the branch. The terminal SNP gets you the rest of the way.

The terminal SNP is the thing, and it seems to me important to use the shorthand in the correct way, rather than monkeying with it to suit this or that person's personal preference. We already have too many different ways of naming things in genetic genealogy.

So what is the terminal SNP going to be for me and for others that are downstream from DF13 but "under investigation" like L526, L580, and L583. Will FTDNA be calling me R-DF13 or R-L583?

That's a good question. My guess is R-DF13.

Unfortunately, this will probably be only as good as FTDNA's ability to keep their programming current.  So, really there is no change there unless they've rewritten some of their stuff and their support processes. Right now, it takes over a year to update their haplotree so a year from now DF13 could seem rather ancient.

We do need to pay tribute to the volunteers at ISOGG who keep their tree current. No programming is required to support applications, such as what a testing company needs, but at least ISOGG keeps the charts current.

Like all good standards efforts, it needs the cooperation of all the stakeholders involved.
Imagine the chaos if we had to use the old "imperial" system of measures instead of the enlightened metric system.:).

"FTDNA last updated their Y haplogroup tree in 2010 and 23andMe in 2011.*  So, they are going by the subclade names that were recognized at those times. In contrast, the ISOGG Haplogroup Tree has been updated over 60 times JUST THIS YEAR! Every time a new SNP is discovered that is upstream of a known SNP (which is happening faster and faster all the time), it has to be inserted into the tree, thus changing the subclade naming pattern. This is why it is so much simpler to just learn the Terminal SNP label."

[rms2]

I didn't mean to sound like a horse's behind about the shorthand earlier. If someone wants to use the R1b-SNP configuration, it doesn't hurt anything and is easy to understand. I just like sticking with the method that appears to be agreed-upon by Hammer and the other scientists. I also like the reminder that R is the major y haplogroup branch in our case.

[hamish]

As DF Reynolds has pointed out, Geno 2.0 testing is likely to change the tree so much that R1b will soon be archaic.

Personally my preferred form of designation would be Y-M343-DF63 :)

[Mikewww]

As DF Reynolds has pointed out, Geno 2.0 testing is likely to change the tree so much that R1b will soon be archaic....

ISOGG's tree is probably as current if not more current than Geno 2.0 already, in terms of new SNPs.  Am I missing something? I recognize there will be some SNPs identified in more than one haplogroup that we don't need know about, but why are we expecting the known Y DNA tree to change because of Geno 2.0, at least the R1b parts of it?

[Richard Rocca]

As DF Reynolds has pointed out, Geno 2.0 testing is likely to change the tree so much that R1b will soon be archaic....

ISOGG's tree is probably as current if not more current than Geno 2.0 already, in terms of new SNPs.  Am I missing something? I recognize there will be some SNPs identified in more than one haplogroup that we don't need know about, but why are we expecting the known Y DNA tree to change because of Geno 2.0, at least the R1b parts of it?

Just speculating out loud here - since National Geographic is heavily involved in genetics projects of tribes with little or no representation in academic studies, it could be that the entire tree from haplogroup A on down will change. Heck, it could be that even the 'R' gets bumped to 'S' or something.

[Heber]

As DF Reynolds has pointed out, Geno 2.0 testing is likely to change the tree so much that R1b will soon be archaic....

ISOGG's tree is probably as current if not more current than Geno 2.0 already, in terms of new SNPs.  Am I missing something? I recognize there will be some SNPs identified in more than one haplogroup that we don't need know about, but why are we expecting the known Y DNA tree to change because of Geno 2.0, at least the R1b parts of it?

Mike,

Here is a screen shot from the Tyler Smith presentation from another forum.
As you can see the relative size of R1b to (for example) R1a is huge.
Even based on the current ISOGG tree the difference is very  large, factor of five.
This tree is based on a sample of 523 individuals from the 1000 Genome project.
Can you imagine what the tree will look like when we get thousands or hundred of thousands or millions on the Geno 2.0 or NGS. My understanding is that one of the goals of Geno 2.0 is to quickly identify new SNPs as the user base increases and add them to the chip for subsequent releases.
Many age estimates are being made by counting SNPs and in this case the expansion of R1b was categorised as extreme.
Another good reason to reform the naming convention.

http://www.imagebanana.com/view/2db2um4f/IMG_8427_CTS_1kg.JPG

http://eng.molgen.org/viewtopic.php?p=4759#p4759

[rms2]

As DF Reynolds has pointed out, Geno 2.0 testing is likely to change the tree so much that R1b will soon be archaic....

ISOGG's tree is probably as current if not more current than Geno 2.0 already, in terms of new SNPs.  Am I missing something? I recognize there will be some SNPs identified in more than one haplogroup that we don't need know about, but why are we expecting the known Y DNA tree to change because of Geno 2.0, at least the R1b parts of it?

Just speculating out loud here - since National Geographic is heavily involved in genetics projects of tribes with little or no representation in academic studies, it could be that the entire tree from haplogroup A on down will change. Heck, it could be that even the 'R' gets bumped to 'S' or something.

Oh, man, I hope that doesn't happen! I'm kind of attached to the R designation. I would really hate to see that change.

[Richard Rocca]

As DF Reynolds has pointed out, Geno 2.0 testing is likely to change the tree so much that R1b will soon be archaic....

ISOGG's tree is probably as current if not more current than Geno 2.0 already, in terms of new SNPs.  Am I missing something? I recognize there will be some SNPs identified in more than one haplogroup that we don't need know about, but why are we expecting the known Y DNA tree to change because of Geno 2.0, at least the R1b parts of it?

Just speculating out loud here - since National Geographic is heavily involved in genetics projects of tribes with little or no representation in academic studies, it could be that the entire tree from haplogroup A on down will change. Heck, it could be that even the 'R' gets bumped to 'S' or something.

Oh, man, I hope that doesn't happen! I'm kind of attached to the R designation. I would really hate to see that change.

Hopefully nobody got a tattoo with the R1b designation! :)

[Terry Barton]

I don't see how using the Terminal SNP by itself will be very useful or helpful.  I think it is going t obe very confusing.  Consider that many Terminal SNPs will be obscure ends of branches (we hope).  And, it's also likely that many Terminal SNPs will be "private".

Take my personal situation - how many Forum readers know where my R-L196 connects to the tree?  (Are you ready to do a search nearly every time you see a kit's Terminal SNP?)

Some of you will probably know where R-L2 is located?  If not, you know U152 (or S28) 

L196 is below U152 and then below L2.  So far, it's found only in Barton Lineage I - though our diversity is sufficient for us to avoid the "private" label.  We are currently R1b1a2a1a1b3c2 at FTDNA and R1b1a2a1a1b2a2 at ISOGG - which is certainly confusing.  (I agree we have to do something)

You would know pretty well where I am on the tree if I listed myself as R-U152, L2, L196 -- or should it be R-S28, L2, L196 ??  (Won't that still depend on who is reporting?)

And - what about the men in my Barton Lineage I.  Most will be reported as a red R-M269, others are green R-P312, R-U152 or R-L2 and then a handful of us are green R-L196.  Won't that freak out the Newbies who know that Haplogroups have to match in order for the ySTR near match to be valid.

I don't see how reporting only Terminal SNPs addresses the needs or issues. 

[gtc]

As DF Reynolds has pointed out, Geno 2.0 testing is likely to change the tree so much that R1b will soon be archaic....

ISOGG's tree is probably as current if not more current than Geno 2.0 already, in terms of new SNPs.  Am I missing something? I recognize there will be some SNPs identified in more than one haplogroup that we don't need know about, but why are we expecting the known Y DNA tree to change because of Geno 2.0, at least the R1b parts of it?

No, you're not missing something. Geno 2.0 is not an SNP discovery trip like WTY is. The Geno 2.0 chip is "loaded up" with known SNPs. The following quote is form Thomas Krahn, a member of the Geno 2.0 design team:

"Also you shouldn't have a too high expectation to find some
ground-breaking new SNP with the Geno 2.0 test. The phylogeny of the
contained markers is pretty much established and only in a few occasions
we will find unexpected parallel or reverse mutations. Geno 2.0 cannot
find absolutely new SNPs, however it will bring high resolution SNP
testing to a very broad user-base."

http://archiver.rootsweb.ancestry.com/th/read/GENEALOGY-DNA/2012-09/1348544973

[osiris]

Maybe certain major SNPs should be assigned as subhaplogroups so you could call them out specifically but leave off the minor ones.  So from R to L2 (where I am) M343, M269 and P312 and assign then maybe B, H and L.  Then I'd be RBHL2a .  But that just looks wrong to me but you'd have some breadcrumbs leading back to R.  Cockamamie I know, but it's hard to squeeze all the data we want in there. 
I kind of think we don't have enough data yet to really figure out the best naming convention yet anyway. Fun to consider though.

[Mikewww]

I don't see how using the Terminal SNP by itself will be very useful or helpful.  I think it is going t obe very confusing.  Consider that many Terminal SNPs will be obscure ends of branches (we hope).  And, it's also likely that many Terminal SNPs will be "private".

Take my personal situation - how many Forum readers know where my R-L196 connects to the tree?  (Are you ready to do a search nearly every time you see a kit's Terminal SNP?)

Some of you will probably know where R-L2 is located?  If not, you know U152 (or S28)  

L196 is below U152 and then below L2.  So far, it's found only in Barton Lineage I - though our diversity is sufficient for us to avoid the "private" label.  We are currently R1b1a2a1a1b3c2 at FTDNA and R1b1a2a1a1b2a2 at ISOGG - which is certainly confusing.  (I agree we have to do something)

You would know pretty well where I am on the tree if I listed myself as R-U152, L2, L196 -- or should it be R-S28, L2, L196 ??  (Won't that still depend on who is reporting?)

And - what about the men in my Barton Lineage I.  Most will be reported as a red R-M269, others are green R-P312, R-U152 or R-L2 and then a handful of us are green R-L196.  Won't that freak out the Newbies who know that Haplogroups have to match in order for the ySTR near match to be valid.

I don't see how reporting only Terminal SNPs addresses the needs or issues.  

Agreed. Let me re-iterate that I'm all for terminal SNP labeling of haplogroups, but for discussions, papers and the like. However, being for something is not a good reason to terminate something else, i.e., the long haplogroup labels.

Alphanumeric phylogenetically meaningful labels are still needed to make it easier, especially for newbies when reading report screens, and for heaven's sake - sorting purposes.

I don't think people recognize the impact this will have on project admins, particularly geographic or surname types who want to sub group by subclade. Do they need to be experts on the complete Y phylogenetic tree? or keep it up and running on a 2nd computer while they subgroup people on the 1st?

This kind of thing (changes to labeling schemes and testing company reporting capabilities) is why I gave up on trying to get to fancy with project screens and decided to dump everything I can into spreadsheets to so I can rapidly sort, select, total or analyze anyway I want. I built a multi-worksheet system where I can import Y DNA SNP reports, compare against a mathematically defined (numbering levels and sub-branches) phylogenetic tree, determine the terminal SNP of any set of test results (within the scope of R1b) and create any label I want to use for sorting, analysis, etc.

... just don't take away the Y DNA SNP report. That's the greatest thing since sliced bread. I just hope we have better ways of getting SNPs from WTY to show up and I most certainly hope Geno 2.0 results will show up in the same SNP labeling system.   I don't need to worry about a new set of labels, rs numbers, to sort through and we don't need a mix of labratory letter style and rs number labeling. Surely they won't try something like that.

[df.reynolds]

As DF Reynolds has pointed out, Geno 2.0 testing is likely to change the tree so much that R1b will soon be archaic....

ISOGG's tree is probably as current if not more current than Geno 2.0 already, in terms of new SNPs.  Am I missing something? I recognize there will be some SNPs identified in more than one haplogroup that we don't need know about, but why are we expecting the known Y DNA tree to change because of Geno 2.0, at least the R1b parts of it?

Mike, we've heard via Thomas and Bennett that Geno 2.0 includes large numbers of SNPs that are currently regarded as equivalent.  My expectation is that with the large number of tests being done, some number of those SNPs will no longer be equivalent.

For example, right now, the ISOGG Hg R tree has 21 SNPs at the R/R1 levels. Thomas' db probably has at least another 20 SNPs that fall around that level. I can easily see R-M343 getting shoved down another level or two in the tree due to new data at the R/R1 level.

--david

P.S. I didn't say R1b would be archaic, I said it might well be metaphorical, meaning that M343 would likely no longer be at the "R1b" position in the tree, but that we would likely to continue to refer the R-M343 as "R1b," even if the clade based name of R-M343 changes.

[gtc]

As DF Reynolds has pointed out, Geno 2.0 testing is likely to change the tree so much that R1b will soon be archaic....

ISOGG's tree is probably as current if not more current than Geno 2.0 already, in terms of new SNPs.  Am I missing something? I recognize there will be some SNPs identified in more than one haplogroup that we don't need know about, but why are we expecting the known Y DNA tree to change because of Geno 2.0, at least the R1b parts of it?

No, you're not missing something. Geno 2.0 is not an SNP discovery trip like WTY is. The Geno 2.0 chip is "loaded up" with known SNPs. The following quote is from Thomas Krahn, a member of the Geno 2.0 design team:

"Also you shouldn't have a too high expectation to find some
ground-breaking new SNP with the Geno 2.0 test. The phylogeny of the
contained markers is pretty much established and only in a few occasions
we will find unexpected parallel or reverse mutations. Geno 2.0 cannot
find absolutely new SNPs, however it will bring high resolution SNP
testing to a very broad user-base."

http://archiver.rootsweb.ancestry.com/th/read/GENEALOGY-DNA/2012-09/1348544973

Further to the above, it strikes me that there may be a point of confusion in  regards to what we might call "new" SNPs versus "known" SNPs.

As Thomas tells it (above), all of the SNPs on the chip are "known" by definition because they have to be -- the chip is programmed to recognize (and report instances of) known SNPs.

However, of those thousands of "known" SNPs on the chip, a subset could be called "new" as far as FTDNA customers are concerned because they haven't been available for testing via FTDNA until they were programmed into this chip because they were only "known" to sources outside of FTDNA and its customer community.

If that's the case, then those of us who have currently tested "to the death" as it were with FTDNA may have the opportunity to be found positive for SNPs that were hitherto unavailable to us via FTDNA, so we could consider those to be new to us (if not to science).