PDA

View Full Version : The Great Divide below DF27



razyn
07-17-2015, 03:34 AM
I've been finding it more and more useful to look at Alex Williamson's Big Tree, and the YFull tree below DF27, to see how much we have learned recently, especially via NextGen sequencing. These two trees are especially agile about revising frequently. I'll try to paste in current links to the parts of these trees that are specific to DF27:
The Big Tree: http://www.ytree.net/DisplayTree.php?blockID=29&star=false
YFull tree: http://www.yfull.com/tree/R-DF27/

Tonight I was noticing what a large number of SNPs now fall under the "leftover" branch -- basically, what we have been calling "DF27+, Z195-" for purposes of (very loosely) organizing subgroups in FTDNA's R1b-DF27 and Subclades haplogroup project. Looking at the Big Tree as displayed for all of DF27, there are 64 discrete branches that lie below Z195/Z196, and 57 that do not. Alex says they are under ZZ12 -- a SNP that is not yet universally acknowledged to exist. (If ZZ12 did not exist, it would be necessary to invent him.)

Anyway, that's getting to be pretty even; and the rate of growth (by discovery) appears to be a good bit faster on the ZZ12 side of this tree. That's a good thing, in that most of the wholly unknown, "asterisked" subclades are still on that side -- the ZZ12 side. The known branches below ZZ12 are populated by the people who have graduated out of the large group Aa, by testing positive/derived for something that isn't under Z195 -- but is (now) identified, and testable. There are still far too many members in group Aa; but even if that list looks about as long as it looked a year ago, they are not all the same people. Many have moved on, but new members have joined the project (after BigY or other tests) to replace them.

More work is steadily being done -- especially (but not exclusively) as many more BigY results come in, and the tools with which we may analyze them are beginning to work a little better. Obviously we also take whatever advantage we can of the results from FGC, YSEQ, and other labs or analysis services. Those clusters, varieties, or even family-specific SNPs that have a strong advocate or voice may be getting more attention, proportionately, than others that may be at least equally deserving, but are less organized. If you feel neglected -- find someone else in group Aa who seems to match you, organize, test -- and pester me.

RCO
07-17-2015, 12:27 PM
Interesting. Why ZZ12 is not included in the YFull tree ?
We can find two types of ZZ12_1 ChrY:19736631..19736631 and ZZ12_2 ChrY:20442843..20442843 in the ISOGG YBrowse
http://ybrowse.org/gb2/gbrowse/chrY/?
What about the Iberian distribution and diversity of branches ?
Thanks

lgmayka
07-17-2015, 02:11 PM
Interesting. Why ZZ12 is not included in the YFull tree ?
We can find two types of ZZ12_1 ChrY:19736631..19736631 and ZZ12_2 ChrY:20442843..20442843 in the ISOGG YBrowse
YF03788 (DF27 > BY653) has:


Sample: #YF03788 (R-P312)
ChrY position: 19736631 (+strand)
Reads: 164
Position data: 97T 67C
Weight for T: 0.581543624161
Weight for C: 0.418456375839
Probability of error: 0.115320099214 (0<->1)
Sample allele: Y (C or T)
Reference (hg19) allele: T



Sample: #YF03788 (R-P312)
ChrY position: 20442843 (+strand)
Sample allele: no call position
Reference (hg19) allele: A

Notice that ZZ12_1 is more often read as ancestral than derived.

But let's say that we treat the minority percentage of derived as a positive (ZZ12_1+). At such a flaky location, it's quite possible that R-DF27 was originally (minority) positive, and Z195+Z196+S355 is the branch that mutated back to pure negative. Isn't it suspicious that right during a period of obviously rapid expansion, DF27 should show a perfect bifurcation (2-way split)--and the allegedly ancestral state hangs on across several SNPs (Z195, Z196, and S355) while the allegedly derived state continues to expand rapidly?

I'm not saying that I know what happened at this location thousands of years ago; but I see no evidence that anyone else really knows for sure either. A SNP needs to be sure and dependable enough for singletons (those who have no shared SNPs more recent than the one in question), and ZZ12_1 does not appear to meet that requirement.

razyn
07-17-2015, 06:20 PM
A SNP needs to be sure and dependable enough for singletons (those who have no shared SNPs more recent than the one in question), and ZZ12_1 does not appear to meet that requirement.

There are purposes to which the term SNP is applied, for which it might require such a doctrinaire definition. But I don't agree that this is one of them.

A UEP (unique event polymorphism) doesn't need to be a SNP (at a single nucleotide), doesn't have to be eternally stable, may be on a palindrome, may occur in a region that is hard to test (or to read, in a test), may be an indel -- and several other things, up with which certain scientists or labs will not put. However: if a UEP clearly indicates a fork in the road, it is IMO kind of silly to ignore the signpost just because it's painted in a tacky color.

That said, I have no particular brief for the stability or longevity of ZZ12, per se. It has a kind of reality, just as guessing the age of a clade by counting the SNPs leading to it (that we know about, and can agree on) has a kind of reality. The notion is useful, even if not immutable, or of a reality that's beyond debate. One may say that the "real" fork under DF27 is at Z195; but that fork also includes Z196, and by your suggestion may include deletion of ZZ12. One may say that... but saying that ZZ12 points at one branch, and Z195 (with equivalents) points at the other lets one draw more elegant diagrams. That also has value.

Rick
07-17-2015, 11:27 PM
where is the ZZ12 test offered?

razyn
07-18-2015, 12:29 AM
where is the ZZ12 test offered?

To the extent that I understand it, it isn't actually "offered;" it has been found as a "rejected" SNP (initially by Alex Williamson, but others know how to find it) in the .bed file, a part of the zipped .vcf file that BigY testers may download, but others can't see. There is no real need to test it, it's just a convenient peg on which to hang the numerous clades and subclades that are DF27+ and Z195-. It's a SNP on a palindrome, those are hard to test and easy to ignore (as a class), so most lab people don't fool with them. It's rejected because the great majority have it only in one direction, say outbound on the palindromic arm but not inbound. One side can copy the other (a RECLOH event); and that is known to have happened at least once when the side with the mutation copied the side without it, and the mutation no longer exists in that lineage (so far, just a branch of the Plant family). They aren't "ancestral" for ZZ12; they have just lost it by another mutation that erased its evidence. It has also happened at least twice when the ZZ12_1 mutation was copied onto the other arm, ZZ12_2, so it tests like a normal SNP -- as derived in either direction. These instances apply to small groups of anciently related families, one such group being Ashkenazi.

This is too inconsistent to be especially useful. But it is evidence of an ancient genetic event, that affected all of the phylogeny below that event -- that is, a signpost to a fork in the road. The ZZ12 road is untraveled by anybody who has the Z195+ mutation, and they are a numerous and pretty well-mapped group. But they may not even be half of the people with DF27 -- we don't have a good census yet. Anyway, the ZZ12 side is clearly quite large, bushy, and old. And that's why I started this thread with "Great Divide" in the title.

Rick
07-18-2015, 01:53 AM
Thanks for the very helpful explanation, Razyn. I wouldn't think a bifurcation would be the necessary state, but the consistent results you report sure present a compelling case. Though, there could be a few DF27*s hiding out there still.

Rick
07-18-2015, 02:19 AM
So are there any maps or tables showing the distributions of the two groups? If not, any general impressions?

razyn
07-18-2015, 02:46 AM
Well, I'm thinking about reorganizing the DF27 project to reflect it, and the ZZ12 side has a tree structure just as surely as does the Z195 side -- it's just far, far less well known. Much of it has only been discovered since FGC and FTDNA started doing NextGen tests, that reveal much more than chip tests, such as Geno2 or Chromo2 (based on SNPS we already knew -- so they are more about sorting than discovery, as such).

At the moment the project groups are alphabetical, and all the early letters were assigned to the SNPs we already knew about in 2013. Many of the new discoveries, alphabetized as they became known, need to be grouped with their brother and cousin clades that already had earlier letters assigned. That's a technically simple, but tedious, job that I find even easier to put off than mowing the lawn. Which is pretty easy.

Earl Davis
07-18-2015, 10:47 AM
Isn't it suspicious that right during a period of obviously rapid expansion, DF27 should show a perfect bifurcation (2-way split)--and the allegedly ancestral state hangs on across several SNPs (Z195, Z196, and S355) while the allegedly derived state continues to expand rapidly?


It's early days and more tests might yet reveal a third branch but Z195 itself seems to have resulted in a potential bifurcation into Z198 and Z272?

Earl.

Earl Davis
07-18-2015, 10:54 AM
So are there any maps or tables showing the distributions of the two groups? If not, any general impressions?

There are over 260 samples now at ytree...

http://www.ytree.net/DisplayTree.php?blockID=29

It still seems to early to draw any conclusions regarding distribution. At this point it's clear that DF27 has a broad distribution with hot spots as diverse as Armenia and Iberia.

razyn
07-18-2015, 03:05 PM
It's early days and more tests might yet reveal a third branch but Z195 itself seems to have resulted in a potential bifurcation into Z198 and Z272?


Or a trifurcation, with DF17 thrown in (or perhaps out). I don't know how many times one may furcate, before it's just not nice. Anyway, it isn't a linear process; but it does resemble a tree (or a not very carefully pruned bush), in that there's a DF27 trunk, somewhere -- back to which every twig has a genetic trail that is (considered in isolation) linear.

And I'd like for our project organization to reflect that, a little better than it currently does. Tweaking it will look strange, especially at the moment one gets tweaked. But it's a housekeeping task that is needed.

Webb
07-18-2015, 10:51 PM
I've been finding it more and more useful to look at Alex Williamson's Big Tree, and the YFull tree below DF27, to see how much we have learned recently, especially via NextGen sequencing. These two trees are especially agile about revising frequently. I'll try to paste in current links to the parts of these trees that are specific to DF27:
The Big Tree: http://www.ytree.net/DisplayTree.php?blockID=29&star=false
YFull tree: http://www.yfull.com/tree/R-DF27/

Tonight I was noticing what a large number of SNPs now fall under the "leftover" branch -- basically, what we have been calling "DF27+, Z195-" for purposes of (very loosely) organizing subgroups in FTDNA's R1b-DF27 and Subclades haplogroup project. Looking at the Big Tree as displayed for all of DF27, there are 64 discrete branches that lie below Z195/Z196, and 57 that do not. Alex says they are under ZZ12 -- a SNP that is not yet universally acknowledged to exist. (If ZZ12 did not exist, it would be necessary to invent him.)

Anyway, that's getting to be pretty even; and the rate of growth (by discovery) appears to be a good bit faster on the ZZ12 side of this tree. That's a good thing, in that most of the wholly unknown, "asterisked" subclades are still on that side -- the ZZ12 side. The known branches below ZZ12 are populated by the people who have graduated out of the large group Aa, by testing positive/derived for something that isn't under Z195 -- but is (now) identified, and testable. There are still far too many members in group Aa; but even if that list looks about as long as it looked a year ago, they are not all the same people. Many have moved on, but new members have joined the project (after BigY or other tests) to replace them.

More work is steadily being done -- especially (but not exclusively) as many more BigY results come in, and the tools with which we may analyze them are beginning to work a little better. Obviously we also take whatever advantage we can of the results from FGC, YSEQ, and other labs or analysis services. Those clusters, varieties, or even family-specific SNPs that have a strong advocate or voice may be getting more attention, proportionately, than others that may be at least equally deserving, but are less organized. If you feel neglected -- find someone else in group Aa who seems to match you, organize, test -- and pester me.

I find using Alex's tree and the Yfull tree side by side very interesting because of the tree structure and Yfull's dates. A tree shows shared and branching snp's much clearer than a spreadsheet when used to connect the dots. I have been staring at the left hand branches of CTS4065 on Alex's tree. I find the Pletcher, Italian, Cilley, Meister, and two Zenkers extremely interesting. A German at the beginning and a cluster of Germans at the end with the Brit and Italian in the middle.

Kvenlander
07-24-2015, 05:07 PM
I was grouped by Alex under ZZ12 and am just learning the whole DF27. But if I understand correctly, ZZ12 would seem to present a sizeable portion of DF27? When looking at the names in this group, there seem to be two big groups, English names and German names. With a quick glance I can only find few Iberian names.

razyn
07-24-2015, 05:37 PM
I was grouped by Alex under ZZ12 and am just learning the whole DF27. But if I understand correctly, ZZ12 would seem to present a sizeable portion of DF27? When looking at the names in this group, there seem to be two big groups, English names and German names. With a quick glance I can only find few Iberian names.

ZZ12 should be everybody in DF27 who isn't Z195+. There is no cheap way to test it; in BigY .bed files it's a "rejected" call on one side of a palindrome, and you can't easily test that on a chip, so they don't want to fool with it. One L617+ family (known so far) should be ZZ12+ but isn't, presumably having had a second mutation at the same site (in the past 3000 years or so) and lost it entirely. Everybody else who is DF27+ and Z195- has it. Some groups have two copies of it (on both sides of the palindrome, one side having copied the other in a RECLOH event).

About the lack of Iberian names, vs. their reported abundance in the literature -- the literature is based on testing that was (a) not at FTDNA, (b) not NextGen, and (c) not done at very high resolution. But if you look at all those pink samples in Alex's Big Tree (that begin with HG0-something), they are from the 1000 Genomes project; and a whole lot of them are "Iberian," at least by one or two removes (sampled in Mexico, Puerto Rico, Colombia, etc.).

REWM
07-26-2015, 09:09 PM
razyn in Alex Williamson's Big Tree I am zz12+.
How can one tell if not derived for ZZ12_1 or zz12_2?


ZZ12_1
ChrY position: 19736631 (+strand)
Reads: 170
Position data: 111T 59C
Weight for T: 0.645751633987
Weight for C: 0.354248366013
Probability of error: 0.206123937522 (0<->1)
Sample allele: Y (C or T)
Reference (hg19) allele: T

ZZ12_2
ChrY position: 20442843 (+strand)
Reads: 1
Position data: 1A
Weight for A: 1.0
Probability of error: 0.0 (0<->1)
Sample allele: A
Reference (hg19) allele: A

razyn
07-26-2015, 10:07 PM
I'm not the person making those decisions, Alex is. Offhand I'd say, if you have 59 calls for the derived state, you probably have the mutation, regardless of the 111 calls for ancestral. (I imagine a Z195+ person wouldn't have any derived calls at all.) But I'm not one of the nerds reading BAM files and making these calls. I'm a nerd about several other things, but not this.

gotten
07-27-2015, 10:08 AM
To support Richard: If you use the chromosome browser on the DF-27 group at Yfull you see that all the people that are under Z195+ get the ancestral state T for ZZ12_1 with high confidence. The people that are Z195- seem to get a 50/50 mix of ancestral and derived. It's a very clean cut between the two groups and so the ZZ12_1 sort of acts like a SNP despite the low confidence in what the state really is. ZZ12_2 seems to be the ancestral value (with a low number of reads) across the whole of DF27. Use the positions 19736631, 20442843, 17922066 if you want to try it yourself.

REWM
07-27-2015, 08:07 PM
Thanks gotten. I did look at YFull DF-27 group already for my z2573 group. One of the 4 of my Z2573 group has only the ancestral value T at position 1973663 and ancestral value A at 20442843. Myself and the other two has mixed reads with more to the ancestral value T. I was not finding fault with razyn. Only trying to understand how it is determined with the mixed reads.

gotten
07-27-2015, 08:24 PM
If all people had the same mixed reads it would be ambiguous and ZZ12_1 would be unusable but in this case there is a clear difference between the two groups. Exactly why you get mixed reads in one case and not in the other... well... we should ask the professionals !

In your case the one person (HG01353) with the ancestral value T at the first position appears to have a very low number of reads at that position. The T is not very significant.

Mikewww
08-03-2015, 04:23 PM
This topic is made more important by the nature of DF27. It is NOT easily tested for. We now know that even though FTDNA has a Sanger Sequencing/Advanced tests option for DF27, as does YSEQ, FTDNA's SNP Pack testing method can NOT read DF27.

I'm an R1b project admin as well as a DF27 admin. BTW, my DF27 lineage is surnamed Pletcher and is Swiss-German.

As an R1b project admin I requested that FTDNA develop an R1b-M343 Backbone SNP Pack. They did build this Backbone Pack but could not include DF27. I tried to include all of the SNPs immediately downstream of DF27 as a work-around. It would be much better if some combination of ZZ12_1 and ZZ12_2 would work.

All of this does not affect the Z195 people as Z195 is safely tested. It's the rest of DF27 that this pertains to.

Another BTW - Dick (Razyn) has done a great job of digging through all of these things for DF27. Thank you. Also thanks to Alex Williamson for expanding the Big Tree to DF27. It's a beautiful thing.:)

REWM
08-03-2015, 06:44 PM
Mikewww I see that Z2573 is included in the backbone test.

Do you know if FTDNA will offer a stand alone test for Z2573? I did the Big Y and it did not test for it, nor did it test for it in my Big Y matches test.
All three of the Z2573+ men on ybrowse had to test it at YSEQ along with several other downstream SNPs of Z2573 to see that we are positive for them. Big Y only found Z29624/FGC32092 which they call a Novel Variant. If someone does test with Big Y they are not be able to test with the backbone test nor as a stand alone.

swid
08-04-2015, 04:08 PM
I figured this is as good as a place to post this question, since R1b panels are a hot topic at the moment; also, it's a decent test case for "generalized advice for anyone who has tested as far as DF27+ only".

My maternal grandfather's line (FTDNA 141266) doesn't have any close matches outside of his own surname, and I couldn't eyeball it as falling into a known R1b STR cluster. A few months back, my curiosity won out and I ordered a few individual high-level SNPs and got him down to DF27; I added him to the DF27 Project at that time. (If anyone wants to entertain guesses as to where this might fall within DF27, be my guest.)

Should I:
- wait for the DF27-specific FTDNA panel to come out? (Pro: should cover all relevant SNPs, presumably. Con: doesn't yet exist.)
- order the R1b backbone panel? (Pro: currently exists for a reasonable price. Con: doesn't test *all* relevant SNPs, so additional testing may be necessary.)
- order the DF27 panel from YSEQ? (Pro: currently exists for a reasonable price. Con: not updated to test *all* relevant SNPs at this moment.)

Mikewww
08-04-2015, 04:59 PM
I figured this is as good as a place to post this question, since R1b panels are a hot topic at the moment; also, it's a decent test case for "generalized advice for anyone who has tested as far as DF27+ only".

My maternal grandfather's line (FTDNA 141266) doesn't have any close matches outside of his own surname, and I couldn't eyeball it as falling into a known R1b STR cluster. A few months back, my curiosity won out and I ordered a few individual high-level SNPs and got him down to DF27; I added him to the DF27 Project at that time. (If anyone wants to entertain guesses as to where this might fall within DF27, be my guest.)

Should I:
- wait for the DF27-specific FTDNA panel to come out? (Pro: should cover all relevant SNPs, presumably. Con: doesn't yet exist.)
- order the R1b backbone panel? (Pro: currently exists for a reasonable price. Con: doesn't test *all* relevant SNPs, so additional testing may be necessary.)
- order the DF27 panel from YSEQ? (Pro: currently exists for a reasonable price. Con: not updated to test *all* relevant SNPs at this moment.)
I posted this on the DF27 yahoo group but my recommendation is to wait for the DF27 specific pack from FTDNA. I definitely would not get the R1b-M343 Backbone Pack. You are beyond that with your Z210 status. The DF27 Pack has been in the works for some time and DF27 has gotten a ton of attention recently. I think you will see a very robust pack. That will be its advantage. A lot more SNPs, the opportunity to break phylogenetic equivalent blocks, etc. It should also result in your haplogroup label being updated on all FTDNA projects that you are in.

The YSEQ panel is quite fine too. It has the advantage of being Sanger Sequenced based. However, I'd still wait. Here's why. On multiple other occasions where FTDNA has come up with SNP Pack, YSEQ has sharpened its game and lowered its price.

Mikewww
08-04-2015, 05:05 PM
Mikewww I see that Z2573 is included in the backbone test.

Do you know if FTDNA will offer a stand alone test for Z2573? I did the Big Y and it did not test for it, nor did it test for it in my Big Y matches test.
All three of the Z2573+ men on ybrowse had to test it at YSEQ along with several other downstream SNPs of Z2573 to see that we are positive for them. Big Y only found Z29624/FGC32092 which they call a Novel Variant. If someone does test with Big Y they are not be able to test with the backbone test nor as a stand alone.
I don't know. I need to check on that see what I can find out.

razyn
08-04-2015, 05:09 PM
My maternal grandfather's line (FTDNA 141266) doesn't have any close matches outside of his own surname

He does, however, have very good matches within it (1 off at 67, and 3 off at 67). If you have made email contact (or already know them), you could chip in $200 each and draw straws for which of the three gets a BigY test. Then he will match somebody (probably outside the surname project) who actually knows what his shared "terminal" SNP is.


Should I:
- wait for the DF27-specific FTDNA panel to come out? (Pro: should cover all relevant SNPs, presumably. Con: doesn't yet exist.)
- order the R1b backbone panel? (Pro: currently exists for a reasonable price. Con: doesn't test *all* relevant SNPs, so additional testing may be necessary.)
- order the DF27 panel from YSEQ? (Pro: currently exists for a reasonable price. Con: not updated to test *all* relevant SNPs at this moment.)

If you want to know soon, and in high resolution, figure out a way to spend what that takes. Either a BigY; or many guessed-at SNP tests; or several "panel" and "pack" tests... none will be cheap, unless you have freakishly good luck.

If you aren't in a big hurry, some more economical routes to a pretty similar level of understanding (for most people, including most DF27+ and Z195- people) will exist in the near future. Maybe six months to a year -- I'm just guessing. The faster we move to rush a DF27 panel test to market, the less completely it will test for newly identified DF27 subclades. (Which btw are still being discovered, about weekly.) But maybe we have found most of the oldest DF27 branches, by now. One hopes so.

Mikewww
08-04-2015, 05:14 PM
He does, however, have very good matches within it (1 off at 67, and 3 off at 67). If you have made email contact (or already know them), you could chip in $200 each and draw straws for which of the three gets a BigY test. Then he will match somebody (probably outside the surname project) who actually knows what their shared "terminal" SNP is....
I forgot to mention NGS testing, but Razyn is absolutely correct. Testing a fixed SNP pack is testing for someone else's SNPs, but with Big Y, you find your own lineage of SNPs. It really is the way to go if you can figure out how to fund it.

swid
08-04-2015, 05:59 PM
I should have prefaced my question with "I know NGS testing is the *most correct* answer to this, but if I decide to do something cheaper...".

The pessimist in me fears that any NGS testing for that line will be a repeat of my own line's experience, and he'd be stuck at a high-level paragroup. Of course, there's only one way to find out... :)

I've been in email contact with a couple of his close surname matches; unfortunately, we've never been able to figure out the exact genealogical relationship, as all of our lines dry up in (different) colonial American locations. I could check to see if any of them are interested in splitting an NGS test, but I've never been very comfortable with asking other people to spend money. I'd make a terrible salesman. :)

miiser
08-04-2015, 06:38 PM
I should have prefaced my question with "I know NGS testing is the *most correct* answer to this, but if I decide to do something cheaper...".

The pessimist in me fears that any NGS testing for that line will be a repeat of my own line's experience, and he'd be stuck at a high-level paragroup. Of course, there's only one way to find out... :)

Your insight serves you well, young Skywalker.

The reality of NGS testing is that if you don't have any close STR matches, then you're unlikely to have any close Big Y matches. Because of this, many Big Y testers end up as singleton branches within a high level "paragroup", having a long list of private SNPs that no one else (or only their close relatives) have. Genealogically relevant information will be gained from NGS testing ONLY if you have matches of interest within a relevant time frame. Whether this is the case can be assessed pretty reliably by looking at your STR matches. Unless there is some unclear phylogeny within your matches due to ambiguous STR signatures, and revealing the exact structure of that phylogeny is critical to understanding the history of your ancestry (for example, is your closest match Irish or Scottish?), then discovering your own branch's SNPs amounts to an expensive novelty item. Unlike yourself, many of the people in these forums are fanatic data enthusiasts who have no qualms about asking other people to spend money, and the advice tends toward, "Do every available test. Money is no object. Feel the power of the dark side."

Trust your instincts, but do not let them guide your actions. If your goal is to learn genealogically relevant information, then there's a good chance Big Y won't add anything beyond what's evident from your STR matches. If you are fascinated by the minutia of knowing your own personal SNPs, have at it.

If it was me, I'd wait for the DF27 panel to come out at FTDNA, and then do either the FTDNA or yseq panel, depending on which one has the more complete collection of SNPs. On the other hand, if your STRs have a strong match to a specific subclade, I'd take a chance and go ahead and test that as a single SNP. In the case of strong STR signatures, jumping straight to the end of the line is the most efficient route. I took a quick look at your kit, and I don't see an obvious match to a specific subclade. So I'd wait for the DF27 panel. If you don't have the patience to wait, then I'd do the yseq panel, followed by single SNP testing as appropriate.

R.Rocca
08-04-2015, 07:43 PM
Your insight serves you well, young Skywalker.

The reality of NGS testing is, if you don't have any close STR matches, then you're unlikely to have any close Big Y matches. Because of this, many Big Y testers end up as singleton branches within a high level "paragroup", having a long list of private SNPs that no one else (or only their close relatives) have. Genealogically relevant information will be gained from NGS testing ONLY if you have matches of interest within a relevant time frame. Whether this is the case can be assessed pretty reliably by looking at your STR matches. Unless there is some unclear phylogeny within your matches due to ambiguous STR signatures, and revealing the exact structure of that phylogeny is critical to understanding the history of your ancestry (for example, is your closest match Irish or Scottish?), then discovering your own branch's SNPs amounts to an expensive novelty item. Most of the people in these forums are data enthusiasts, and the attitude tends to be, "Test everything. Money is no object. Feel the power of the dark side."

Trust your instincts, but do not let them guide your actions. If your goal is to learn genealogically relevant information, then there's a good chance Big Y won't add anything beyond what's evident from your STR matches. If you are fascinated by the minutia of knowing your own personal SNPs, have at it.

If it was me, I'd wait for the DF27 panel to come out at FTDNA, and then do either the FTDNA or yseq panel, depending on which one has the more complete collection of SNPs. On the other hand, if your STRs have a strong match to a specific subclade, I'd take a chance and go ahead and test that as a single SNP. In the case of strong STR signatures, jumping straight to the end of the line is the most efficient route. I took a quick look at your kit, and I don't see an obvious match to a specific subclade. So I'd wait for the DF27 panel.

I think this is sound advice, since you answered someone who is already a couple of levels below L21. However, for the masses, there are some pitfalls to the approach you've outlined...

1. We are starting to see instances where customers are going straight to NGS and bypassing STR testing altogether.
2. For the great majority of folks that are close to modal haplotypes even at 37 or 67 markers, NGS testing will tell you quite a bit more than just STRs. Case in point, all of my closest matches on Y-Search are all false matches, even all of the ones that are U152 and L2 which are four levels up from my current subclade.
3. If nobody did NGS testing, then we would be at an SNP standoff. NGS is not just about matching SNPs with someone already in FTDNA's database, but also with future testers.

swid
08-04-2015, 08:06 PM
Just so that anyone doesn't get confused in future replies, while my own Y-line is below L21, my maternal grandfather's line is DF27, hence the discussion in this thread.

I was using the experience from my own testing as an example of a "worst-case" potential outcome for going the NGS route with his line. I've had the luck of dealing with two Y-lines that, even in the ridiculously well-tested Isles, seem to have no (yet) known Y-line relatives much closer than the initial R1b explosion. :)

I know that it's a chicken-and-egg thing *and* that you never know what you'll learn until you test. I'm trying to balance my thirst for data of any sort with trying to do things in a cost-effective manner and balancing all the tradeoffs that entails.

razyn
08-04-2015, 08:22 PM
And btw the confusion is that ZZ10 (swid's SNP below L21) has been misread as Z210 and applied, erroneously, to his maternal grandfather's line. That worthy has tested DF27+ (but nothing further downstream, so far). He has 67 markers, so that's not a problem. He and his excellent 67-marker matches are all called M269 (in red) by FTDNA; but in fact are almost certainly all some shade of DF27 -- as one has tested -- and all share a surname, with minor spelling variation. Almost everybody who takes the BigY test now is getting some kind of match, and I have suggested that 141266 Hayes is likely to follow that recently achieved pattern.

Earl Davis
08-06-2015, 11:31 AM
Mike mentioned elsewhere that ZZ12 is now on the FNDNA R1b panel. If it works (fingers crossed) and because L617 is also on the panel so the ZZ12 reLOC guys will be covered off as well then the R1b panel should be a lot more useful to the 'other half' of DF27 than it originally appeared.

If ZZ12 does work this is where these types of SNP pay off for me. Even if it's a little less cut and dried than some other SNPs a positive result for a R1b person will provide important clues as to what more stable snips to test for rather than being stuck with a basic P312.

Regards Earl.

Earl Davis
08-06-2015, 11:45 AM
I did some very basic analysis of the of the occurrence of DF27 subclades yesterday. I intend to take this down to the fourth level under DF27 eventually with DF27 being level 0 for this purpose. It's all pretty imperfect at the moment once you hit level two due to the low volume of samples and the variance in snips in each block but for what it's worth...

This is based on a sample of 280 DF27 results on Alexs' Big Tree as of 5th August.

Level 0
Lead SNP # Samples % of DF27 LVL0
DF27 280 100%

Level 1
Lead SNP % of DF27 LVL0
Z195 137 49%
ZZ12 143 51%

Level 2
Lead SNP % of DF27 LVL0
Z198 59 21%
Z272 78 28%
FGC20747 12 4%
Z225 13 5%
DF83 10 4%
BY653 3 1%
Z2552 14 5%
26286073-C-T 2 1%
Y6951 3 1%
FGC23066 3 1%
ZZ19 54 19%
DF79 4 1%
Z2573 4 1%
Undiscovered 17 6%

razyn
08-25-2015, 06:33 AM
Since I basically believe the Great Divide is real, and I was running out of alphabet for creating a new subgroup every time somebody got a new BigY match, I moved a lot of groups around (by giving them a different letter, color, description or whatever). Basically, to agree a little better with the Big Tree. It might be viewed as a work in progress, especially the colors. Was also running out of rainbow -- FTDNA only purports to offer 38 colors, some of which don't work or default to a similar one, and several aren't in the rainbow. Anyway, there was a little method to it. Part of it, of course, was just madness. The SNP names may or may not agree with what someone else calls at the same position, etc.

The former group Aa is now group E, since everybody in that group (DF27+, Z195-) will basically end up in a group farther down the alphabet. This way, you don't have to scroll past all those greens and yellows to arrive at some SNP you might possibly want to test. I also tried to eliminate some unnecessary repetition of the statement that these groups are all R1b, P312, DF27. It is, after all, the DF27 project. I haven't yet done much new on the Z195+ side of the Great Divide; it's messy, but at least most of the basic haplogroups are sorted within B (and, in the separate and older "SRY2627" project, for C).

razyn
08-28-2015, 04:54 AM
I haven't yet done much new on the Z195+ side of the Great Divide; it's messy, but at least most of the basic haplogroups are sorted within B (and, in the separate and older "SRY2627" project, for C).
I neglected to mention (because I hadn't done it yet) that former groups C and D have swapped identifying letters. C now identifies DF17 (which after all belongs next to the B group); and D identifies Z198, the group formerly identified as L176.2. Z198 and its haplogroups are still yellow, but DF17 is now "Lime Green."

In a further attempt to rationalize the DF27 project's groups, I have redefined the former group Aaa (now Ea) to include those persons who have had a BigY test (or other NextGen sequencing of the Y), but have not yet found matches by which a new "terminal" SNP might be confirmed and a subgroup defined. This is closer to the real meaning of DF27**, as better testing rapidly expands our view of the full extent of DF27. The group consists initially of nine people I could find (most of them in group E) who met those criteria. Seven of them are already at the far right on the Big Tree. I will request better data uploads from the other two. There may well be more members who are so tested, especially if it wasn't done at FTDNA. Feel free to point them out to me.

I also created a new group Sb (no known kin to Sa; just running out of letters for tiny new groups). This SNP CTS5329 is on the Big Tree, with our intermittently visible member 193923 and an anonymous sample from the 1kG project. We actually have another BigY match, but he's not yet a project member, so I can't display him in the new "group" yet. It's kind of like a tree in winter; you just need to believe there will be leaves.

razyn
01-08-2016, 07:02 PM
This thread is in need of updating in the light of the new SNP packs. As it turns out, both ZZ12 and ZZ19 are successfully tested with the new DF27 SNP pack, which isn't actually a chip test -- it uses the same sequencing machinery as BigY, but with an array of very much smaller targets. It's above my pay grade to explain it, but I did read a blog post about the explanation that was given at the November FTDNA administrators meeting.

On the ZZ12+ (or "Z195-") side of this thread's Great Divide, the biggest subgroup is that under ZZ19. When I reorganized the project in August, all of the ZZ19 subclades began with the letter F and some number, or number plus lower case letter. This quickly became unwieldy, the pace of discovery having already outrun my ability to subdivide clades using that system. But it is better than nothing.

Anyway, under ZZ19, most of the subclades (found so far) fall below another important SNP, Z31644. It comes at the end of a string of "problematic" SNPs that are hard to test, expensive to test, hard to believe in, ambiguous reads, and so on: DF27>ZZ12>ZZ19>Z31644. Below it (ergo, below that whole string) one finds A431, and all the other groups from F1 through F7. Within the last couple of months a handful of BigY tested samples have shown positive for Z31644, but with no further shared SNP identified below that. When there were only a couple of them, I tacked them on at the end of the list (F7a); but really they are the asterisk group, "basal" Z31644* until proven otherwise. I suspect that eventually they will be so proven, but for the time being I have created a group F -- mainly, to move them to the beginning rather than the end of the list of Z31644's numerous subclades.

This creates an appearance of inconsistency in naming group F8, with its especially large Rox2 component; those are subclades of ZZ19, but not of its major Z31644 branch. I did at least give them a darker shade of olive, back when I set up these F subgroups. If you care about the phylogenetic detail, just bear in mind that the color changes probably indicate something significant; and either the caption or the Big Tree should indicate what that is. The alphanumerical codes may, or may not. In several cases they just say, "and here are some other guys."

R.Rocca
01-29-2016, 02:52 PM
Does anyone have a single master list of all of the SNP positions that have been found positive/derived in DF27 men?

ArmandoR1b
01-29-2016, 03:17 PM
If someone has that I would like to have a copy of it also.

Mikewww
01-29-2016, 06:26 PM
Does anyone have a single master list of all of the SNP positions that have been found positive/derived in DF27 men?
The Big Tree is the best representation of all NGS derived SNPs for DF27 people. Only shared SNPs are shown, but he might have the novel SNPs under the covers from his web site.
I've downloaded most of them for review and inclusion (or not) in DF27 SNP Packs. I have that in a spreadsheet format and do an ISOGG like row by row tree except phylogenetic equivalents get their own lines.

How are you trying to use this? Maybe my spreadsheet format will work. I have, from time to time I've also scanned (and added) the YFull tree and Ray Banks' composite tree but this an eye test and mistake prone.

razyn
01-29-2016, 06:42 PM
YSEQ, YFull, FGC, Chris Morley, Jim Wilson and others have their own lists. I doubt if there are any two lists of the SNPs below DF27 currently extant that agree with each other, let alone include everything. And people write academic papers without so much as a glance at any of the more complete ones.

We're working on it. It still changes more or less weekly, having slowed a bit from several time a week.

R.Rocca
01-29-2016, 07:03 PM
The Big Tree is the best representation of all NGS derived SNPs for DF27 people. Only shared SNPs are shown, but he might have the novel SNPs under the covers from his web site.
I've downloaded most of them for review and inclusion (or not) in DF27 SNP Packs. I have that in a spreadsheet format and do an ISOGG like row by row tree except phylogenetic equivalents get their own lines.

How are you trying to use this? Maybe my spreadsheet format will work. I have, from time to time I've also scanned (and added) the YFull tree and Ray Banks' composite tree but this an eye test and mistake prone.

Going over ancient DNA samples to see if there are any SNPs that may match DF27/DF19/DF99 below just the first couple of branches. So yes, a spreadsheet format would be ideal.

Mikewww
02-01-2016, 11:43 PM
Going over ancient DNA samples to see if there are any SNPs that may match DF27/DF19/DF99 below just the first couple of branches. So yes, a spreadsheet format would be ideal.
I have to go three speadsheets but I will combine them tonight and send you the best that I have. Basically it is a combination of Williamson's Big Tree, Yfull tree and administrator highlighted Chromo 2 SNPs.
I will work DF27 first.

Mikewww
02-02-2016, 02:49 PM
I have to go three speadsheets but I will combine them tonight and send you the best that I have. Basically it is a combination of Williamson's Big Tree, Yfull tree and administrator highlighted Chromo 2 SNPs.
I will work DF27 first.
You should should have it by now.

If you didn't get it I must have the wrong email ID for you, so please email me at m w w d n a at gmail.

R.Rocca
02-02-2016, 03:13 PM
You should should have it by now.

If you didn't get it I must have the wrong email ID for you, so please email me at m w w d n a at gmail.

Thanks Mike, I will see if there is anything I can find.

razyn
02-02-2016, 04:04 PM
Thanks Mike, I will see if there is anything I can find.
I just sent Mike a list of twenty new ones (below DF27>ZZ12>FGC20747) this morning. I was writing it up while you guys were chatting over here.

As I said before, a "complete" list is just about impossible. It changes with most incoming NextGen test results; and they don't come in to everybody at the same pace, level of completeness, and so on. Most of the changes now aren't at a high level on the tree, but one never knows what will turn up next.

Good luck with the aDNA data, anyway.

Connell Warrior
02-14-2016, 02:50 PM
My Y Full initial results are in and they say I am R-FGC14124.

lgmayka
02-14-2016, 08:08 PM
My Y Full initial results are in and they say I am R-FGC14124.
But your new entry may have forced the creation of a new subclade. Wait for the next version of the R-FGC14124 haplotree (http://yfull.com/tree/R-FGC14124/).

Here's a hint: In your YFull account, go to the "Hg and SNPs" page. At or near the top of the list, are there any SNPs marked "terminal"? If so, these may define a new subclade in the next version of the tree.

Connell Warrior
03-09-2016, 08:03 PM
It did Igmayka, they have now listed me with Mike Durkin as R-M428! Also saying based on our samples 950ybp!

razyn
03-28-2017, 02:50 PM
Maciamo Hay reorganized his Eupedia chart of DF27 last month, and it now reflects the Great Divide to which my original July, 2015 post (this thread) refers. Alex Williamson's Big Tree has reflected it all along; that's actually where I got the impression that there was such a divide. But it's even more pronounced, a year and a half later, as visually reorganized by Maciamo. I'll paste in a screen shot, but in order to grab the whole thing in one screen it has to be in a very small font. So if you want to look at it full size and actually read it, here's the url at which you scroll down until you find DF27 (twice): http://www.eupedia.com/europe/Haplogroup_R1b_Y-DNA.shtml#R1b-subclades

And here's the screen shot:

14735

MitchellSince1893
03-28-2017, 03:07 PM
Maciamo Hay reorganized his Eupedia chart of DF27 last month, and it now reflects the Great Divide to which my original July, 2015 post (this thread) refers. Alex Williamson's Big Tree has reflected it all along; that's actually where I got the impression that there was such a divide. But it's even more pronounced, a year and a half later, as visually reorganized by Maciamo. I'll paste in a screen shot, but in order to grab the whole thing in one screen it has to be in a very small font. So if you want to look at it full size and actually read it, here's the url at which you scroll down until you find DF27 (twice): http://www.eupedia.com/europe/Haplogroup_R1b_Y-DNA.shtml#R1b-subclades

And here's the screen shot:

14735

Solely based on Maciamo's labels both branches appear to be widespread from the Western to Eastern ends of Europe, with one from Kyrgyzstan in Central Asia.

Earl Davis
03-28-2017, 06:48 PM
It's good to see the progress made in ZZ12 over the last 3 years is finally starting to be reflected across the internet

Taking a look back at July 2013 pretty much nothing was known about ZZ12's subclades when the then 'known' D27 tree looked like

https://s-media-cache-ak0.pinimg.com/736x/59/02/eb/5902ebeb02588f1ca78c7c95a3f3535c.jpg

and it seemed that Z195/Z196 was dominant.

It's not quite 4 years later and Z195 now appears to be the smaller branch of the two accounting for perhaps 40-45% of DF27.

It's nice to see more of these diagrams finally showing how widespread DF27 is across most of Europe and West Asia. It was a hard argument to make only a couple of years ago when many papers were still focussed on Iberia and the Basques.

Earl.

Earl Davis
03-28-2017, 07:11 PM
It's also worth noting that it's almost 4 years since we started to order our original batch of BigY tests. Hopefully the incoming 'Chromium technology' tests will enable us to make the next leap forward in the granularity we need to better define genealogically meaningful branching of ZZ12 in the years ahead. With luck continued advances could see BigY as an entry level introductory test in a few years time. This coming from someone who remembers when a Y12 STR test seemed like the cutting edge of genetic genealogy testing.

Earl.

lgmayka
03-28-2017, 10:21 PM
Solely based on Maciamo's labels both branches appear to be widespread from the Western to Eastern ends of Europe, with one from Kyrgyzstan in Central Asia.
Even individual clades are sometimes widely spread. The diagram shows BY653 in Britain and Ukraine; YFull now also shows it in Sardinia (https://yfull.com/tree/R-BY653/).

Webb
03-29-2017, 10:59 AM
Even individual clades are sometimes widely spread. The diagram shows BY653 in Britain and Ukraine; YFull now also shows it in Sardinia (https://yfull.com/tree/R-BY653/).

And with a TMRCA of 4200 ybp, we are still a long ways away from figuring out this small branch of DF27, and really a large number of similar branches which are small and old and widespread.

R.Rocca
03-29-2017, 12:33 PM
For those that look at DF27 religiously, how is this "great divide" divided geographically?

Webb
03-29-2017, 01:10 PM
For those that look at DF27 religiously, how is this "great divide" divided geographically?

There is no rhyme or reason to the geographical spread. The Z195 side of the divide has been tested longer, so this side has quite a few more snp stacks or runs. But if you look at the picture that Razyn linked, you will see there is not much of a difference, geographically between the two large clusters under DF27. There is coverage from Scandinavia to Eastern Europe, to Southern Europe, to Western Europe.

R.Rocca
03-29-2017, 01:48 PM
There is no rhyme or reason to the geographical spread. The Z195 side of the divide has been tested longer, so this side has quite a few more snp stacks or runs. But if you look at the picture that Razyn linked, you will see there is not much of a difference, geographically between the two large clusters under DF27. There is coverage from Scandinavia to Eastern Europe, to Southern Europe, to Western Europe.

When we throw out the one-off far flung samples that may be the result of relatively recent movements, does that change anything?

Webb
03-29-2017, 03:08 PM
When we throw out the one-off far flung samples that may be the result of relatively recent movements, does that change anything?

I am not a big fan of removing the outliers, personally. An example is L165. Jean M. and many others have closed the book on this marker being Scandinavian. Its presence in the Scottish Islands has seemed to confirm it. However, there has been one longtime French sample that has been largely ignored. Now there is a new French sample on the Big Tree and the English samples have grown. So after several years, the Scandinavian samples are now the outliers.

razyn
03-29-2017, 05:13 PM
When we throw out the one-off far flung samples that may be the result of relatively recent movements, does that change anything?
It smooths the way for the ever-popular oversimplification. ("DF27 the Iberian haplogroup," or whatever.) But "wide is the gate, and broad is the way, that leadeth to destruction." Matt. 7:13, KJV. That's not a theological turning point, it's just folk wisdom, and good advice.

Far-flung examples (like A431 in an Armenian valley) may be the result of relatively recent movements. Or they may not be. Most of DF27 is yet to be tested to levels at which the laws of large numbers might kick in, and make phylogeographical deduction from the statistics reliable. Right now, for the most part we still can only speculate.

But we have a boatload more to speculate with than we did in 2013, when the separate DF27 project was started -- at your suggestion, btw. See the last sentence in the following post, #11 on its thread, in which you were talking to me. And my response was in post #20, same thread: http://www.anthrogenica.com/showthread.php?942-Z195-and-SRY2627-in-Flanders-New-paper&p=6660&viewfull=1#post6660

Part of the newer, more complete picture is this "great divide." Another part is considering U152 and DF27 as brothers below ZZ11, more distantly kin to L21 (and all that other stuff, often beginning with ZZ, that Alex has been teasing out of higher resolution sequencing). Another part is widening the base map (eastward). All of these steps have their ripple effects, but it takes a while.

R.Rocca
03-29-2017, 05:30 PM
It smooths the way for the ever-popular oversimplification. ("DF27 the Iberian haplogroup," or whatever.) But "wide is the gate, and broad is the way, that leadeth to destruction." Matt. 7:13, KJV. That's not a theological turning point, it's just folk wisdom, and good advice.

Far-flung examples (like A431 in an Armenian valley) may be the result of relatively recent movements. Or they may not be. Most of DF27 is yet to be tested to levels at which the laws of large numbers might kick in, and make phylogeographical deduction from the statistics reliable. Right now, for the most part we still can only speculate.

But we have a boatload more to speculate with than we did in 2013, when the separate DF27 project was started -- at your suggestion, btw. See the last sentence in the following post, #11 on its thread, in which you were talking to me. And my response was in post #20, same thread: http://www.anthrogenica.com/showthread.php?942-Z195-and-SRY2627-in-Flanders-New-paper&p=6660&viewfull=1#post6660

Part of the newer, more complete picture is this "great divide." Another part is considering U152 and DF27 as brothers below ZZ11, more distantly kin to L21 (and all that other stuff, often beginning with ZZ, that Alex has been teasing out of higher resolution sequencing). Another part is widening the base map (eastward). All of these steps have their ripple effects, but it takes a while.

Without putting geographical labels on anything, is there an obvious bifurcation areas between ZZ12 and its sublcaldes and Z195 and its subclades? For example, within U152's largest subclade L2, there are some obvious east/west patterns around the Rhine.

MitchellSince1893
03-29-2017, 07:14 PM
Without putting geographical labels on anything, is there an obvious bifurcation areas between ZZ12 and its subclades and Z195 and its subclades? For example, within U152's largest subclade L2, there are some obvious east/west patterns around the Rhine.

And on some subclades, 20+ SNPs below L2 (as seen on ytree.net) there are still branches with samples found throughout Western and Central Europe.

In these cases one may need to be 25-30 SNPs below DF27 and U152 (roughly 1st millennium BC) before you start seeing some geographic clusters

razyn
03-29-2017, 07:23 PM
"Obvious east/west patterns around the Rhine" would still be geographical labels, in my book. Anyway, I think the "great divide" itself is so near basal DF27 that its geography isn't likely to be very visible. A little clan of pretty mobile guys (pastoralists, horsemen, boatmen or something) would perhaps stay together, initially, even over a wide swath of Eurasia. But a step or two down, it matters a lot -- and the sons/grandsons etc. of Mr. Z195 don't have the same patterns of settlement/diffusion that the sons/grandsons of ZZ12 have -- even though those guys were "brothers." Maciamo's pair of descent charts (viewed at full size, so one may read the geographical captions) now hint at these more divergent aspects of DF27 more realistically than his previously well-known heat maps do.

The little flag symbols on the Big Tree also carry hints; but so many of those MDKA flags are in the western hemisphere, they don't really tell us what we'd prefer to be told. Which basically is why we are all waiting for the aDNA papers. Theoretically there is no DF27 aDNA in Brazil, Puerto Rico, Colombia etc. (modern countries prominent in the Big Tree flag display).

GoldenHind
03-29-2017, 11:02 PM
The little flag symbols on the Big Tree also carry hints; but so many of those MDKA flags are in the western hemisphere, they don't really tell us what we'd prefer to be told. Which basically is why we are all waiting for the aDNA papers. Theoretically there is no DF27 aDNA in Brazil, Puerto Rico, Colombia etc. (modern countries prominent in the Big Tree flag display).

I have noticed that Alex is inconsistent in his use of the little flags, so I wouldn't place too much emphasis on them.

R.Rocca
03-30-2017, 12:43 PM
"Obvious east/west patterns around the Rhine" would still be geographical labels, in my book. Anyway, I think the "great divide" itself is so near basal DF27 that its geography isn't likely to be very visible. A little clan of pretty mobile guys (pastoralists, horsemen, boatmen or something) would perhaps stay together, initially, even over a wide swath of Eurasia. But a step or two down, it matters a lot -- and the sons/grandsons etc. of Mr. Z195 don't have the same patterns of settlement/diffusion that the sons/grandsons of ZZ12 have -- even though those guys were "brothers." Maciamo's pair of descent charts (viewed at full size, so one may read the geographical captions) now hint at these more divergent aspects of DF27 more realistically than his previously well-known heat maps do.

That's what I'm trying to understand... what is that differentiating pattern a couple of steps down that differentiates between the sons/grandsons of Mr. Z195 and the sons/grandsons of ZZ12?


The little flag symbols on the Big Tree also carry hints; but so many of those MDKA flags are in the western hemisphere, they don't really tell us what we'd prefer to be told. Which basically is why we are all waiting for the aDNA papers. Theoretically there is no DF27 aDNA in Brazil, Puerto Rico, Colombia etc. (modern countries prominent in the Big Tree flag display).

We can only guess that most of the Brazilian, Puerto Rican and Colombian flags represent Iberian colonialists, but those flags without corresponding surnames are meaningless. For example, SE Brazil had very large Italian and German immigration during the early 20th century.

razyn
03-30-2017, 04:00 PM
That's what I'm trying to understand... what is that differentiating pattern a couple of steps down that differentiates between the sons/grandsons of Mr. Z195 and the sons/grandsons of ZZ12?
Well, again, I think you're sort of grasping at straws, given the limited data. But I suppose I could point to a couple of trends.

On the Z195 side, in the Z295+ lineages, there's a pretty clear division between what I have previously called the CTS4065 branches (that probably should be backed up to S25783), and what I may have called CTS12074 (but Z216, Z270 or other tags at that higher level might work better). The former is turning up almost (but not quite) exclusively on the NE side of the Pyrenees (but then all over Europe except for Iberia). The latter has a very pronounced Iberian flavor, only lightly sprinkled with several English surnames. And it's the side that includes (about ten rungs down the ladder) the "Basque marker" M153. So that looks enough like a pattern, with enough data, probably to have meant something, pretty anciently.

On the ZZ12 side, some major subclades have a strongly if not exclusively Iberian flavor. I think this is most visible on the Big Tree, where the pink background (indicating incompletely analyzed data) very often refers to 1000 Genomes samples from, shall we say, Iberian America? Anyway one can tell that data source from the NA or HG kit numbers, whether the little flags are present or not. This broad generalization would tend to be evident (so far) in ZZ39, ZZ51 (though DF83 has British subclades), FGC20747 (with both Iberian and Jewish ancestry prominent, often in the same sample) and Z2552. Moving to the right (as displayed currently by Alex) the ZZ19/ZZ20 group looks much less skewed toward Iberia. And again, even in the most markedly lopsided larger haplogroups, there may be a subclade or two that contradict the rule.

Also, there are Dutch, Norwegian, Scottish and other "clusters," perhaps originally discovered via STR markers, but now associated with a SNP or group of SNPs. Some of those are below Z195, and others below ZZ12. They are in several cases old enough to be reflecting ancient movements, at least pre-Roman (and not the post-Roman Crusaders, Visigoths, Napoleon's army and so on). I don't think the presence or absence of a subclade in Iberia is its only significant characteristic; it's just one that gets discussed a lot, because of a pre-existing condition we have (being tagged as an Iberian haplogroup).

And another thing we can't see is the presumed large number of ancient lines that daughtered out, and aren't being tested in the modern population.

ehjelt
04-09-2017, 03:25 PM
YFull Ytree 5.03 is here but where is The Great Divide?

Mikewww
04-09-2017, 04:00 PM
YFull Ytree 5.03 is here but where is The Great Divide?
I am not at my computer but I think that ZZ12 is not in a CombBED region and YFull will not use it. Their approach seems to focus on reliability in defining branches and not granularity.

razyn
04-09-2017, 04:04 PM
YFull has criteria that exclude several categories of SNPs. ZZ12 is in one of the several such categories. (It's on a palindrome, so one side can copy the other in a RecLOH event -- so it may have disappeared, or doubled itself, in some lineages over several millennia. We know of a couple of cases in which either of these has happened, sometime since the ZZ12 mutation distinguished itself from the Z195+ side of the tree a few thousand years ago. But we, meaning I personally, don't care; ZZ12 is a handy project-organizing tool, and was real enough in the Bronze Age.) So their tree is based on well-sequenced and highly detailed data (but for BigY results, only what gets submitted to them, with $49 -- by no means all the BigY data present in FTDNA's bank). The same is true for several other testing companies' sequenced data. Like ISOGG, and most other tree-displaying organizations, YFull then makes decisions about which discoveries, from the vast data available to them, they will deign to include in their tree. Their perspective on palindromes excludes ZZ12. So their tree just has a large and heavily populated branch for Z195+ -- no corresponding SNP that marks the rest of the DF27 tree, as known so far. This works for them, but it looks strange to people I have parked in subgroup E of the DF27 project. In a sense, that's my way of saying to those members, "Forget about the green and yellow stuff above you in this display."

razyn
05-03-2017, 12:31 AM
Looking at the Big Tree as displayed for all of DF27, there are 64 discrete branches that lie below Z195/Z196, and 57 that do not. Alex says they are under ZZ12 -- a SNP that is not yet universally acknowledged to exist. (If ZZ12 did not exist, it would be necessary to invent him.)

Anyway, that's getting to be pretty even; and the rate of growth (by discovery) appears to be a good bit faster on the ZZ12 side of this tree.

I've shamelessly quoted myself, once again -- in this case, from the post with which I opened this "Great Divide" thread. Nearly two years later, my hinted prediction that the ZZ12 side would overtake the Z195 side has proven accurate.

Looking at Alex Williamson's Big Tree as of this date (May 2, 2017), there are 273 samples assigned to 133 discrete branches (subclades) on the Z195+ side of the tree -- which in 2015 was much better understood than the Z195- side (aka ZZ12+). On the ZZ12 side of Alex's tree there are 390 samples assigned to 191 discrete branches. Measuring the two groups either by branches (133/191) or by samples (273/390), the Z195 group is 7/10 the size of the ZZ12 group.

That may not be enough to justify a paper in [I]Nature; but we've been digging these yams for several years now, and one pile has steadily been getting bigger, compared with the other. That means something. So I just mention it -- again.

MitchellSince1893
05-03-2017, 02:49 AM
I've shamelessly quoted myself, once again -- in this case, from the post with which I opened this "Great Divide" thread. Nearly two years later, my hinted prediction that the ZZ12 side would overtake the Z195 side has proven accurate.

Looking at Alex Williamson's Big Tree as of this date (May 2, 2017), there are 273 samples assigned to 133 discrete branches (subclades) on the Z195+ side of the tree -- which in 2015 was much better understood than the Z195- side (aka ZZ12+). On the ZZ12 side of Alex's tree there are 390 samples assigned to 191 discrete branches. Measuring the two groups either by branches (133/191) or by samples (273/390), the Z195 group is 7/10 the size of the ZZ12 group.

That may not be enough to justify a paper in [I]Nature; but we've been digging these yams for several years now, and one pile has steadily been getting bigger, compared with the other. That means something. So I just mention it -- again.

ZZ12 is now roughly the same size as all of U152 combined on ytree.net. 390 vs 389 individuals and 191 vs 195 branches.

Watch out L21. DF27 coming after you next. :)

ADW_1981
05-03-2017, 02:17 PM
There is a P312+ kit from Egypt in the P312+ Atlantic project. Wonder if he's DF27?

R.Rocca
05-03-2017, 02:59 PM
ZZ12 is now roughly the same size as all of U152 combined on ytree.net. 390 vs 389 individuals and 191 vs 195 branches.

Watch out L21. DF27 coming after you next. :)

In had predicted this a few years back given that DF27 is about twice as frequent in England as U152 and given the very pronounced increase of Americans of Spanish ancestry (via Latin America) that are testing with FTDNA. The more data the merrier.

MitchellSince1893
05-03-2017, 06:30 PM
In had predicted this a few years back given that DF27 is about twice as frequent in England as U152 and given the very pronounced increase of Americans of Spanish ancestry (via Latin America) that are testing with FTDNA. The more data the merrier.

All those Irish L21s immigrants to America in the 1800s and Ulster Scots/Scotch-Irish in the 1700s, will help keep L21 in the lead for quite a while, but I think eventually the Hispanic American testers may help DF27 catch up to L21 in the FTDNA database and Alex's tree.

Mikewww
05-04-2017, 01:01 AM
All those Irish L21s immigrants to America in the 1800s and Ulster Scots/Scotch-Irish in the 1700s, will help keep L21 in the lead for quite a while, but I think eventually the Hispanic American testers may help DF27 catch up to L21 in the FTDNA database and Alex's tree.
I agree. There has to be ton of DF27 in Latin America.

razyn
05-04-2017, 04:15 AM
Comment removed also.

MitchellSince1893
05-04-2017, 05:10 AM
Response removed.

Webb
05-04-2017, 01:16 PM
Latin America can easily be predicted as can the British Isles. What I want is France and Germany. More testing. More results.