PDA

View Full Version : R1b-L21 Phylogeny (SNP based family tree)



Mikewww
04-16-2013, 05:50 PM
I try to keep this graphic current. It is the Y DNA tree/phylogeny for L21 as best as I know it. Please post any corrections or ommissions here.
https://dl.dropboxusercontent.com/u/17907527/R1b-L21_Descendency_Tree.jpg

FTDNA's haplotree is way, way behind so for source data for the graphic I use International Society of Genetic Genealogy's (ISOGG) Y hapgroup R tree at
http://isogg.org/tree/ISOGG_HapgrpR.html

I supplement that with the fine research David Reynold's does for us on his Walk The Y (WTY) project and Geno 2.0 (from National Geneographic Project) comparative tables such as these at
http://daver.info/geno/results/R-L21.pdf
http://daver.info/geno/results/R-DF21.pdf
http://daver.info/geno/results/R-DF49.pdf
http://daver.info/geno/results/R-L513.pdf
http://daver.info/geno/results/R-L1335.pdf
http://daver.info/geno/results/R-Z253.pdf

David also assembled this comparison of data in the 1000 Human Genome project.
http://daver.info/1000Genomes/HgR-L21.htm

Another source of data is FTDNA's lab director Thomas Krahn and the draft tree he posts for L21.
http://ytree.ftdna.com/index.php?name=Draft&parent=99813537

Of course, another source is new news from our R1b-L21 project Y DNS SNP report screen. I check it about once a week.
http://www.familytreedna.com/public/R-L21/default.aspx?section=ysnp

rms2
04-18-2013, 06:18 PM
That is really a great graphic and extremely useful when trying to explain where things stand now, SNP-wise.

Mikewww
04-26-2013, 09:34 PM
Z251 appears to be a solid, diverse peer with the larger subclades of DF13 so I added it as a rectangle. There are couple of unknowns about its relationship with smaller DF13 subclades.
https://dl.dropboxusercontent.com/u/17907527/R1b-L21_Descendency_Tree.jpg

History-of-Things
05-12-2013, 05:17 AM
Regarding the positioning of L1336: I always forget exactly what the clouds mean. To my knowledge L720 (if I am to be taken as an example) is L1336- but the place it is put looks to be upstream of us--maybe reconsider how to position it? Thanks. I had to think about that one a second, since I'm always ready to help out with new testing to clarify things, but the WTY already indicates L1336-. Lacking a little excitement these days with nothing new and useful to test....

danmac21
05-16-2013, 02:14 PM
Thanks for the tree. It is very helpful in orienting to this topic.

Mikewww
06-03-2013, 12:49 PM
Here are the latest updates. Among other things I've inserted F4036 and PF825 as independent peers under Z253 but it is unclear whether they are independent of each other, plus F4036 may be private. The grayed out SNPs are ones that have very high probabilities of remaining in private status. That is a bit of subjective.

I've also got another URL that leads back to the same long dropbox URL. This should be easy to remember -
http://tinyurl.com/R1b-L21-Tree



Putting an SNP in a "cloud" just means its positioning is not completely clear, it's still a little cloudy or it means it is private. Since ISOGG will not track private SNPs their positioning will always be subject to cloudiness because there is no official cross-checking of all the relevant SNPs against it.

Mikewww
06-21-2013, 12:00 AM
L21 will be updated on the ISOGG tree soon.
Only David would apologize for all of the volunteer work he is doing. L21ers are lucky to have him.

My apologies, but I am still travelling, and still behind on my e-mail and other
responses. If you awaiting a response from me (particularly for something posted
here), please feel free to e-mail me again and/or directly (dna (at) daver.info)
to ensure I don't miss responding to you.

I've sent in a number of edits to Alice, and we should see the changes made to
the ISOGG tree within the next couple weeks at the latest. Following SNPs are
being added under the R-L21 tree:

R1b1a2a1a2c1f3-PF825.2
R1b1a2a1a2c1g2a1a-L658 (moved)
R1b1a2a1a2c1g2a1b-CTS3655
R1b1a2a1a2c1g2a1b1-L627 (moved)
R1b1a2a1a2c1g2a1c-L1402
R1b1a2a1a2c1g5-L1336
R1b1a2a1a2c1k1a-L743
R1b1a2a1a2c1l-CTS4466

Waiting for a CTS4466+ Z251 result before adding Z251 to the tree.

I am sneaking ahead and adding what I think Z251 will be.

http://tinyurl.com/R1b-L21-Tree

Mikewww
07-02-2013, 09:10 PM
Z251 was added to the ISOGG tree. I also noted citizen-scientist Andy G's early reports on DF85 & DF97 (correction). Sounds significant to me, and the positioning is clear so I'm adding them to the chart.



Discovery and validation of a new branch below M222

As some of you know we have been working with public genome data for some time. I am pleased to report the identification of a new branch below M222. We have identified two novel variants shared by samples from the 1,000 genomes project and the personal genome project, namely NA20318 and hu2E413D.

The shared variants are defined as:
DF85, a C>A mutation at position 23334448
DF97, a G>A mutation at position 22631679

Testing anonymous samples, I have determined that DF85 is above DF97 in the phylogenetic tree. i.e we have men who are DF85+ DF97- but not the other way around.

-M222
--DF85
---DF97

Since the samples are anonymous I have no indication which M222 men may be positive for these markers. All I know for sure is that I am negative (ysearch 7QEWC)

Primer details have been forwarded to the relevant consumer testing companies. http://archiver.rootsweb.ancestry.com/th/read/DNA-R1B1C7/2013-07/1372767795

http://tinyurl.com/R1b-L21-Tree

razyn
07-02-2013, 09:42 PM
Z251 was added to the ISOGG tree. I also noted citizen-scientist Andy G's early reports on DF83 & DF97. Sounds significant to me, and the positioning is clear so I'm adding them to the chart.
Should be DF85 -- DF83 is one of ours.

Mikewww
07-02-2013, 10:15 PM
Should be DF85 .
Got it. Thank you for checking.

Mikewww
08-19-2013, 09:43 PM
CTS2457.2 and L679 have been added by ISOGG as additional subclades of DF13:

• • • • • • • • • • • • R1b1a2a1a2c1m CTS2457.2
• • • • • • • • • • • • R1b1a2a1a2c1n L679

I also think CTS3386 is on its way to public status. There are GDs of 8 at 67 with only two known haplotypes. The other two CTS3386+ guys are Geno 2 people without Y STRs.
I think David Reynold's fine work has already shown CTS3386 is ancestral in the other subclades of DF13. I don't see anything missing at this point.
http://daver.info/geno/results/R-L21.pdf
... probably just need another set of Y STRs or two among the z4371418 variety which look to be CTS3386+.

http://tinyurl.com/R1b-L21-Tree

Mikewww
10-07-2013, 02:41 AM
I saw a note today on the M222 forum that says Jim Wilson has a tree for downstream of M222. I'm try to draw it from the narrative. This is my understanding of the L21 tree with the latest news on M222 subclades.

Essentially, I think that S661 and S568 are two peers under M222. S661 has S588 and S673. S673 appears to be equivalent to DF85. We already think DF97 is downstream of DF85.

Under S568 is S555 through S567.

Let me know if you have different information.

http://tinyurl.com/R1b-L21-Tree
http://archiver.rootsweb.ancestry.com/th/read/DNA-R1B1C7/2013-10/1381103318

David
10-07-2013, 03:58 AM
I saw a note today on the M222 forum that says Jim Wilson has a tree for downstream of M222. I'm try to draw it from the narrative. This is my understanding of the L21 tree with the latest news on M222 subclades.

Essentially, I think that S661 and S568 are two peers under M222. S661 has S588 and S673. S673 appears to be equivalent to DF85. We already think DF97 is downstream of DF85.

Under S568 is S555 through S567.

Let me know if you have different information.

http://archiver.rootsweb.ancestry.com/th/read/DNA-R1B1C7/2013-10/1381103318

Looks correct as far as it goes. Note there were also few SNPs added under R-DF21, in the Little Scots Cluster.

http://www.anthrogenica.com/showthread.php?170-DF21-(L21-gt-DF13-gt-DF21)-and-Subclades-(DF5-S191-P314-2-S190-etc)&p=15403&viewfull=1#post15403

--david

Mikewww
10-07-2013, 01:15 PM
Looks correct as far as it goes. Note there were also few SNPs added under R-DF21, in the Little Scots Cluster.

http://www.anthrogenica.com/showthread.php?170-DF21-(L21-gt-DF13-gt-DF21)-and-Subclades-(DF5-S191-P314-2-S190-etc)&p=15403&viewfull=1#post15403

--david

I just updated the chart by adding S425 underneath S190. I'll have to update the spreadsheet to support the other SNPs as equivalent.

Looks like I'm about to run out of chart space for an all in one L21 chart.

P.S. What's cool about dropbox and this forum is that I just save the updated chart in the dropbox folder and the way the forum link to picture URL function works when the forum screen is refreshed (by the user) the updated charts is there. The dropbox link for this graphic sits behind http://tinyurl.com/R1b-L21-Tree I've got the same setup for DF27, P312 and L513.

Mikewww
11-15-2013, 04:14 PM
Progress on discovering the public R1b-L21 phylogenetic tree has been incremental but steady over the last couple of years.

This is an almost current descendancy tree: http://tinyurl.com/R1b-L21-Tree

Besides the master L21 haplogroup project, there are almost twenty L21 subclade projects. We can now count over 2300 haplotypes of 111 STRs.

I'm missing a few, but looking back I can count.
63 WTY's completed
391 Geno 2.0's completed

However, progress is about to accelerate to a new level.
131 Chromo 2's on order or completed
117 Big Y's on order
27 FG's on order or completed

With the rush of Big Y's, the major subclades of L21 are now all covered, although we'd like to see several diverse tests per subclade. Chromo 2 seems to be especially beneficial to M222 people along with a major divider, DF85.

Here is even better news for L21. On Sunday, Nov. 17th, National Genographic will unveil results for 100 Co. Mayo Irish folks.
http://www.mayo.ie/dnn/NewsSports/GenographicProject.aspx#.UoX1QvmD3Tp
I would guess the majority of those will be L21+.

Over the next six to nine months or so should see the unveiling of the true public L21 phylogenetic tree.

IrishTypeIII
11-30-2013, 07:40 AM
When I look at http://daver.info/geno/results/R-Z253.pdf I see Kit # 74219 has tested L226+.
As I am interested in all L226+ men I would like to see his STR results. 74219 does not appear in either the Z253 or the L21 projects. Is this kit number correct? Or should it be 72419 McIntyre?

Dennis

Mikewww
11-30-2013, 04:15 PM
When I look at http://daver.info/geno/results/R-Z253.pdf I see Kit # 74219 has tested L226+.
As I am interested in all L226+ men I would like to see his STR results. 74219 does not appear in either the Z253 or the L21 projects. Is this kit number correct? Or should it be 72419 McIntyre?

Dennis

I think it is a typo. This is one of the reasons I copy/paste stuff into a spreadsheet. I have an internal index so if a kit # is not found in the index it's easy to see.

Mikewww
12-02-2013, 05:01 PM
My graphic tree has started to become out of date as Chromo 2 results start coming in. I don't know if a Chris Morley type tool will be gathering data from Chromo 2 as well.

It's time for me to come up with a new way of tracking SNPs for L21 folks that integrates in to the R1b-L21_Haplotypes spreadsheet. Right now I handle S series SNPs manually and that won't work in the future and I'm not sure how Big Y and FG will impact this. I will have to separate the SNPs spreadsheet work from the STRs spreadsheet work internally (for myself) so that will add another step to merge them back together. However, I think I'll expose the internal Y DNA SNP work I do because I'll need to create a "Y DNA SNP report" for Chromo 2, similar to the FTDNA project report screen.

RobertCasey
12-02-2013, 09:13 PM
Here are my preliminary review of L21 mutations found in just three Full Genomes tests P312/DF99, Z253/L226 and Z253/PF825. This does not even cover novel YSNPs which will add another 50 to 100 new L21 YSNPs:

All SNPs below are negative for P312/DF99:
Approx L21 (positive for both Z253s): M2694, PF5501, Z252
Approx Z2534 (positive for L226, negative for PF825): CTS3524, CTS3525, K389, M6006
Approx PF825 (positive for PF825, negative for L226): CTS10596, CTS12273, L1495.2, L430.2, PF2916

Not sure if these are even stable mutations or possibly known equivalent YSNPs or the same YSNPs by different names. Only two are testable by FTDNA currently: L1495 and L430 (which must
be recurrent since these have tested positive already in different parts of the haplotree). The Full Genome reports do not indicate the YChr position for most of these YSNPs since they have
already been discovered and other known sources of YChr find a few but not most of them. Need somebody to cleanse this list of mutations in unstable areas.

The FGC YSNPs are proving a real challenge since many high quality private YSNPs do not have FGC designations in the original files sent but are being sent out as updated files. Also, many
of the high quality shared novel YSNPs have FGC designations while many others do not. I think that I will have to track them by YChr position first - then have a cross reference for
YChr positions to FGC assignments. I am waiting on updated versions that have many more FGC assignments for the other two submissions (mine is more comprehensive in assignments
since my submission was analyzed much later).

I think that we will be lucky to just get ISOGG YSNPs on the graphic in another month or two. There will probably be around 1,000 new L21 YSNP mutations with just the Full Genomes tests
and the Big Y will surely add another 1,000 more. I guess I will have plenty of "private" YSNPs to analyze for months to come.

Mikewww
12-03-2013, 03:50 PM
....
I think that we will be lucky to just get ISOGG YSNPs on the graphic in another month or two. There will probably be around 1,000 new L21 YSNP mutations with just the Full Genomes tests
and the Big Y will surely add another 1,000 more. I guess I will have plenty of "private" YSNPs to analyze for months to come.

Yes, it's both scary and good at that same time. We need systematic, open (public) and documentable ways to work through this so the analysis could be redone by multiple parties with repeatable results.

Mikewww
12-04-2013, 10:21 PM
Yes, it's both scary and good at that same time. We need systematic, open (public) and documentable ways to work through this so the analysis could be redone by multiple parties with repeatable results.

This is critical for Chromo 2 result sharing and comparisons.

I'm starting the process to integrate Chromo 2 results in the R1b-L21_Haplotypes spreadsheet. This will bring on a large number of new SNPs, phylogenetically equivalent SNPs and different labels for the same SNPs.

You can find this spreadsheet in the Links (under More down arrow) section of the R1b-L21-Project yahoo group. I plan to keep this file name and URL the same.
https://dl.dropboxusercontent.com/u/17907527/R1b-L21_SNPs.zip

Currently this is just a draft spreadsheet and is incomplete. This is actually an extension to an internal spreadsheet tab(worksheet) that you have not seen before.

In this spreadsheet, think of the two tabs, "FTin" and "C2in" as input data storage areas for generally relevant SNP results at the individual haplotype level. "FTin" is really just a copy/paste of the R-L21plus FTDNA project Y DNA SNP report screen.
http://www.familytreedna.com/public/R-L21/default.aspx?section=ysnp

C2In is the new piece and is my attempt to create a shared report for L21 Chromo 2 Genetic Signature results. The data for this will all come from the positive/derived SNP Genetic Signature .txt files stored in the R1b-L21-Project yahoo group Files storage under the Chromo 2 results Folder.
http://groups.yahoo.com/neo/groups/R1b-L21-Project/files/Chromo_2_L21_S145_results

It is imperative for all L21+ people with Chromo 2 results to get their Genetic Signature .txt into this folder.

The (future) power of this spreadsheet will be to allow people to use the column heading autofiltering to look for various combinations of positive/derived SNP results, enabling discovery of the R1b-L21 phylogenetic tree. You don't see this yet, but I will greatly expand the Clade tab/worksheet to include the S series SNPs from Chromo 2. Ultimately, this will allow me to appropriately include the S series SNPs in the large R1b-L21_Haplotypes spreadsheet in the Downstream/Relevant SNPs column. This will also allow for more comprehensive labeling in the Haplogroup column of that spreadsheet.

This is a work in progress. Help! Where I am going off track?

Please double check the UpstreamSNPs tab/worksheet. This is my attempt to list SNPs upstream of L21 in both the FTDNA/Geno 2 world and the BISDNA/Chromo 2 world. I'm probably missing some and/or have some that should not be included. I use these "upstream" SNPs to filter out useless (except to scientists) SNPs that we all have. In other words this is part of the steps of filtering out SNPs ancestral for all of L21. I've got some extra columns I'll probably delete in future versions so as not to bog you down but I'll leave some of those in as I get up to speed.

I'll eventually add a tab/worksheet for all of my old WTY/23andMe reported/manual results but this is what I'm trying to move a way from. Manual processes introduce errors and are very inefficient.

I'm kind of wait and see on the FG and Big Y results. There may be so many derived SNP results from them other kinds of filtering have to take place. I'm hoping Mark J will be able to considate relevant FG results (and it would have its own "FGin" tab on the above spreadsheet) and I'll figure out to handle Big Y. FTDNA's promises on providing integrated reporting for Big Y should resolve that but we have to see it first to know if it does or doesn't need additional work.

Essentially what I'm working on so far is making Geno 2 and Chromo 2 more broadly useful for comparative analysis.

Mikewww
12-06-2013, 07:26 PM
I'm starting to run into a problem - overlapping SNP labels. For example, S524 searches end up find CTS5248 as well. Including the + or - resolves that but CTS524 and S524 is even more problematic.

Computers can do a lot but unless we are doing real programming, they need specific identifiers.

BISDNA and Chris Tyler-Smith are causing me problems... the same, to a lesser degree on Paolo Francalacci and the F series SNPs.

I hate to do it, but in my spreadsheet logic I may have to start going with "CT" rather than "CTS" for those SNP labels and "PA" for rather "PF" for those. Does Chris really need three letters? :)

I can see why some people might want to take all of these and translate them into a different naming system that is a little more logical.... next thing you know you'll end up with long haplogroup labels though. oh brother!

greystones22
12-06-2013, 09:47 PM
Mike I have switched to using the map locations eg 13500000 G>A. That's bombproof I think, except for pesky S SNPs.

Mikewww
12-06-2013, 10:58 PM
Andy, any guesses on how many public branches of the R1b-L21 tree we might have six months from now? I'm not sure how young "public" is officially but let's say at least 600 years old.

Mag Uidhir 6
12-07-2013, 01:10 AM
Andy, any guesses on how many public branches of the R1b-L21 tree we might have six months from now? I'm not sure how young "public" is officially but let's say at least 600 years old.
Mike,
Since P66+ is ISOGG & FTDNA accepted and it appears after L69.5 (unstable)...I estimate P66 circa 1400ish maybe as late as 1600. The grouping within L69.5 appears to predate surnames (IMO) with an eponym of Udhir circa 950AD. Believe the SNP occurred shortly before this...but hoping for further SNP discovery within BGY/FGC test in prog to narrow the window.

In short, "public" may be younger than 400?

Brad

Mikewww
12-13-2013, 09:53 PM
Mike,
Since P66+ is ISOGG & FTDNA accepted and it appears after L69.5 (unstable)...I estimate P66 circa 1400ish maybe as late as 1600. The grouping within L69.5 appears to predate surnames (IMO) with an eponym of Udhir circa 950AD. Believe the SNP occurred shortly before this...but hoping for further SNP discovery within BGY/FGC test in prog to narrow the window.

In short, "public" may be younger than 400?

Brad

I guess what's important is what does ISOGG thinks is "public." P66 has been around so long it may have made the tree under less stringent requirements. I think happened with M37. M37 was later demoted. I guess I shouldn't post anything and just my mouth shut on it or the McKown guys won't be happy.

RobertCasey
12-13-2013, 11:27 PM
Mike, I looked at your spreadsheet and I think that we will have to split up (normalize) the L21 spreadsheet to keep the YSNP strings reasonable. Even then, spreadsheets may be not feasible anymore. Let's say that we get 100 Full Genomes tests at 40 new L21 YSNPs each by mid next year (current rate is higher but future ones will share a lot of YSNP mutations) and 1,000 Big Y tests at a very conservative 10 new L21 YSNPs by mid next year. That is 14,000 YSNPs to be listed in the two columns for Full Genomes and Big Y combined. We will need all the negative results to conduct the analysis (working with FG to get this published in the near future). We also will need negatives for the Big Y test (as we did for the Nat Geo uploads via David's spreadsheet). Even we divided up L21 into 10 groups, that is still 1,400 YSNPs per column just by mid-year. Just with the first ten FG tests, you will need 400 YSNPs, then the next 10 tests will require 350 more, etc. I guess we need to be careful what we wish for (a deluge of new YSNPs to analyze).

Mikewww
12-14-2013, 01:58 AM
Mike, I looked at your spreadsheet and I think that we will have to split up (normalize) the L21 spreadsheet to keep the YSNP strings reasonable. Even then, spreadsheets may be not feasible anymore. Let's say that we get 100 Full Genomes tests at 40 new L21 YSNPs each by mid next year (current rate is higher but future ones will share a lot of YSNP mutations) and 1,000 Big Y tests at a very conservative 10 new L21 YSNPs by mid next year. That is 14,000 YSNPs to be listed in the two columns for Full Genomes and Big Y combined. We will need all the negative results to conduct the analysis (working with FG to get this published in the near future). We also will need negatives for the Big Y test (as we did for the Nat Geo uploads via David's spreadsheet). Even we divided up L21 into 10 groups, that is still 1,400 YSNPs per column just by mid-year. Just with the first ten FG tests, you will need 400 YSNPs, then the next 10 tests will require 350 more, etc. I guess we need to be careful what we wish for (a deluge of new YSNPs to analyze).

The spreadsheets I create are not necessarily useful for true discovery and publication of the SNP based tree. They are not "all in one" tools nor fit for true research. They are more for analysis and comparison of STRs and SNPs to determine what SNPs should be investigated more fully. They would not help one fully understanding of ancestral versus no call reads. The kind of analysis that Greg Magoon or David Reynolds (or you or Alex) have done is still fully needed.

This is more for the rest of the folks, the settlers, after the explorers have already gone ahead and created maps. However, settlers can connect genetic genealogy dots. Even settlers stumble upon new discoveries from time to time, at least within their realm.

I recognize the new challenge presented by hundreds or thousands of new data variables that have singular unique rather than a finite number of data variables that have multiple values (like STRs do). There is still the commercial challenge for these new data items (SNPs) with singular (pertinent when derived) unique values. The more recent the derived value the smaller the market place. Now maybe the day comes when everyone gets their true whole genome sequenced for $100. I don't think that is 2014, though. I could be wrong. If I'm not wrong, then the meshing in Y STRs for approximation on the tree and SNPs for the rest is still feasible.

Justin has made a good point on another thread. One at a time testing will be too expensive, therefore a la carte Sanger sequencing is economically problematic. Extreme precision in targeting via the STR values may be needed, at least for a while.

I'm a little bit nervous about which SNPs are truly stable versus fairly recurrent. That's a whole other question. I'm a little worried that many SNPs aren't going to be that useful.

In reality, we need something like watson's capabilities for non-structured data pattern recognition and cognitive computing. That's not in my pocketbook. The net is you end up with character strings and string manipulation functions.

I guess the other alternative is the death of large haplogroup projects. I'm in an SNP boat where I only really care about the last 1000 years. That narrows the scope of relevant SNPs. All of the upstream stuff is of academic importance, rather than genetic genealogy. I'm interested in that too, but the amount people will pay/contribute for that is different.

RobertCasey
12-16-2013, 05:45 PM
I just do not see how the Big Y data will be able to added to the FTDNA reports, as this would create huge strings that make Nat Geo 2.0 uploads look very small. I think at best we will get only the positives reported and will have to deal with negatives like we do for Nat Geo uploads. Nat Geo 2.0 uploads, probably Big Y tests and FTDNA's future haplogroup deep clade tests will only probably only report positive YSNPs in YSNP reports. This may be the solution to reducing the bloat but we can not get ISOGG qualified or determining father/son relationships without negative test results. This would greatly reduce your YSNP list and allow it to remain in spreadsheet form - but it would lose a lot of its usefulness without negative test results present.

Also, we need to make some progress on individual testing of Full Genomes tests via YSEQ - just to keep the ball rolling. I think waiting until dozens of Big Y and FG tests to become available would be way too slow. Plus we need to keep the ISOGG tree growing so that not everyone has to take a $500 test or wait for another $200 static test like Nat Geo 2.0 or the haplogroup deep clade test. I have got four Approx Z253 YSNPs to test via YSEQ - yet no takers yet - very surprising. These are high quality stable YSNPs (one third are probably duplicates but two thirds are probably branch defining branches which around 75 % will be ISOGG qualifying since 75 % of the timeframe is older than 600 years). I am also very concerned that the PF825 test and my L226 test having no common YSNPs between us. The only explanation is that we share very little post Z253 ancestry (less than 5 % or some should have been found).

Highly recurrent YSNPs is now less of problem for new FGC YSNPs since we have a database of over 1000 testers to compare with. Unstable areas may be another issue - but those plus duplicates should be only in the 30 to 40 % ranges and is the price of doing the business of YSNP discovery.

Mikewww
12-16-2013, 08:57 PM
I just do not see how the Big Y data will be able to added to the FTDNA reports, as this would create huge strings that make Nat Geo 2.0 uploads look very small. ...
I agree or I guess I should say I wonder how they will do it, too. I think need to call them and see what they are up to. We have to prepare for it whatever it is.

In a sense, what you were saying about splitting the spreadsheet and what I was saying about the potential death of large haplogroup projects are just reflections of the issue.

From a genetic genealogy perspective, it makes sense for FTDNA to have a reference model for each major subclade of DF13. That's similar in concept also to what I was talking about on other threads about SNP data and "Ysearch for SNPs". On those threads I proposed an idea of having a versioned configuration of relevant SNPs per each major subclade. It's all about the same issues.

In the current spreadsheet I've created a "filters" to screen out upstream of L21 SNPs. I wouldn't hope to try to handle everything upstream of L21 back to a chimpanzee reference model or whatever scientists use. I would think FTDNA would have to apply some kind of concept like that.

There are two approaches. The one I was using in my old version of the spreadsheets of was an exclusive L21 relevant only approach. Everything else was ignored.

However, because of the large number of potentially useful SNPs, I've moved a way to an inclusive everything but what's filtered out approach. As I mentioned, this is in the current spreadsheet and it works okay for Geno 2, Chromo 2 and what we've had in the past from WTY, but has no hope for the results of a FGC output or potentially Big Y output. Essentially, I'm expecting a filtering step for FGC and Big Y before I dump in what looks like is relevant.... however, you know I've asked for your estimate of the number of public SNPs L21. The reason hopefully is apparent. The shifting definition of what's public plays into this... it kind of draws a line in the sand, but that's only semantics anyway.

Sorry for the probable confusion and mumblings with what I'm talking about.:)


... Also, we need to make some progress on individual testing of Full Genomes tests via YSEQ - just to keep the ball rolling. ... .

That's what I'm hoping to do by incorporate promising FGC SNPs into the spreadsheet. That will give us a vehicle for discussion on the R1b L21 yahoo group, whether it is through additional FGC testing or YSEQ. I've got a little more data on that so I'll take a deeper look. For example, I was just looking at the variety I had Mark J assigned in and the Watterson folks. I think there is a good chance that FGC5496 is fairly old. It could be another Z251.

Mikewww
12-17-2013, 11:03 PM
There are many interesting things going on with L21, even without the Next Generation Sequencing (NGS).

Chromo 2 has broken the M222 tree along multiple branches, one of which follows a major split with DF85. We also see more branching upstream between DF23 and DF49.
Chromo 2 has also split CTS4466/Irish II and found branching within L1065/Scotts.

I just noticed that two little known subclades of DF13, CTS3386 and CTS1751, just picked up new STR signatures. Geno 2 is still paying dividends by bring new people from unexpected Y STR signatures.

On NGS, the most exciting news is Mark J's two SNPs, FGC5496, which looks to be old, and further downstream FGC5495. It turns out FGC5495 is identical to S series SNP S7958, that has 11 other folks already marked by it.

Mikewww
12-18-2013, 02:15 PM
... On NGS, the most exciting news is Mark J's two SNPs, FGC5496, which looks to be old, and further downstream FGC5495. It turns out FGC5495 is identical to S series SNP S7958, that has 11 other folks already marked by it.

Mark reminded me that S7958/FGC5495 might be problematic (some kind of x cross-over possibility) so I should be cautious about its abilities to flag a branch on the L21 tree.

Mikewww
12-18-2013, 03:51 PM
I've updated my spreadsheets on the R1b-L21-Project yahoo group. It is intended for members of the R-L21 project who then are eligible to join the yahoo group.

There are now three spreadsheet files:

R1b-L21_Haplotypes
R1b-L21_Haplotypes_Ext
R1b-L21_SNPs

I keep the web addresses (URLs) for these three files constant so you can bookmark them as favorites.

The first one is still very large but I've split the 111 STR marker haplotypes off into their own, smaller spreadsheet. I did this to help for people who wanted a slimmed down version. There are over 2400 111 STR haplotypes so there is still a lot there. The bigger spreadsheet has over 11,000 haplotypes, ranging from 12 to 67 markers.

The SNPs speadsheet is where I keep my understanding of the L21 public tree, consolidate SNPs from multiple sources (i.e. FTDNA, Chromo 2 and FGC), filter out upstream and error-prone SNPs and transform (insert) equivalent SNPs to their designated "lead with" labels as far as haplogroup labeling goes. It might make more sense to go back to the old long haplogroup labels now. Not sure? Probably not.

hoxgi
12-19-2013, 01:29 PM
Mike, thank you for the SNPs spreadsheet, which is very helpful already in analysing new SNPs.

Could you please check it for two SNPs from Chromo 2.0, namely CTS927 and S2906? Of the 12 current Chromo 2.0 results in the L21 Yahoo group, CTS927 and S2906 are positive in the two Z253+ kits, the three L1335+ kits and the one DF49+ kit, whereas your spreadsheet only has four positive results for each when I use the filter.

These two SNPs are negative in the other Chromo 2.0 results (one DF13*, three L513+, one DF21+ and one DF41+) - except for a no-call on CTS927 in the DF21+ kit.

So they appear to define a branch of DF13 which then divides into Z253, L1335 and DF49.

Joss Le Gall's results are inconsistent with the above, as he is Z253+ and negative for these two SNPs, but only his initial results are available at present, not the updated ones.

Greg H

Mikewww
12-19-2013, 01:36 PM
Mike, thank you for the SNPs spreadsheet, which is very helpful already in analysing new SNPs.

Could you please check it for two SNPs from Chromo 2.0, namely CTS927 and S2906? Of the 12 current Chromo 2.0 results in the L21 Yahoo group, CTS927 and S2906 are positive in the two Z253+ kits, the three L1335+ kits and the one DF49+ kit, whereas your spreadsheet only has four positive results for each when I use the filter.

These two SNPs are negative in the other Chromo 2.0 results (one DF13*, three L513+, one DF21+ and one DF41+) - except for a no-call on CTS927 in the DF21+ kit.

So they appear to define a branch of DF13 which then divides into Z253, L1335 and DF49.

Joss Le Gall's results are inconsistent with the above, as he is Z253+ and negative for these two SNPs, but only his initial results are available at present, not the updated ones.

Greg H
That would be cool to find another very ancient layer of branching in L21. I will check it out. The caveat is I've had problems with the CTS and S series of labeling since the letter "S" is embedded at the end of "CTS" in CTS labeled SNPs.

Hopefully I'm looking at the same list. I've got these guys as all CTS297+ S2906+. I don't see any errors so far.

f143812 MacRae R1b-L21>DF13>L1335>L1065>S744>S764
f285203 Cameron R1b-L21>DF13>L1335>L1065>S744>S691>S695
f48641 MacMillan R1b-L21>DF13>L1335>L1065

f200481 Buchanan R1b-L21>DF13>Z253
f288343 Barry R1b-L21>DF13>Z253>Z2534>Z2185>L1066.1

f289737 Willis R1b-L21>DF13 ("*")

z0IDUNB Dunbar R1b-L21>DF13>DF49>S476>DF23>Z2961>M222>S7073>S658>DF104>DF105>S588>S7814+

BTW, Willis is the guy that has the FGC5495+ synonym of Mark J's. I wonder what Mark J's result is at CTS927? We don't have any Y STRs for Dunbar, but notice the deep branching Chromo 2 has for M222?

EDIT/update: I double checked the raw results that people gave us. The spreadsheets look like they are working for this. As long as CTS297 and S2906 are stable and Chromo 2 testing is not error prone on this, it looks like a super clade uniting L1335 (including Scots), Z253 (which is all over W. Europe), DF49 (including the NW Irish) and some DF13* folks. I sent a request to Jim Wilson to see what he says. He needs to look at the latest results for Le Gall but he has quite a number of Chromo 2 folks that we don't have.

Mikewww
12-19-2013, 02:41 PM
That would be cool to find another very ancient layer of branching in L21. I will check it out. The caveat is I've had problems with the CTS and S series of labeling since the letter "S" is embedded at the end of "CTS" in CTS labeled SNPs.

Hopefully I'm looking at the same list. I've got these guys as all CTS297+ S2906+. I don't see any errors so far.

f143812 MacRae R1b-L21>DF13>L1335>L1065>S744>S764
f285203 Cameron R1b-L21>DF13>L1335>L1065>S744>S691>S695
f48641 MacMillan R1b-L21>DF13>L1335>L1065

f200481 Buchanan R1b-L21>DF13>Z253
f288343 Barry R1b-L21>DF13>Z253>Z2534>Z2185>L1066.1

f289737 Willis R1b-L21>DF13 ("*")

z0IDUNB Dunbar R1b-L21>DF13>DF49>S476>DF23>Z2961>M222>S7073>S658>DF104>DF105>S588>S7814+

BTW, Willis is the guy that has the FGC5495+ synonym of Mark J's. I wonder what Mark J's result is at CTS927? We don't have any Y STRs for Dunbar, but notice the deep branching Chromo 2 has for M222?

EDIT/update: I double checked the raw results that people gave us. The spreadsheets look like they are working for this. As long as CTS297 and S2906 are stable and Chromo 2 testing is not error prone on this, it looks like a super clade uniting L1335 (including Scots), Z253 (which is all over W. Europe), DF49 (including the NW Irish) and some DF13* folks. I sent a request to Jim Wilson to see what he says. He needs to look at the latest results for Le Gall but he has quite a number of Chromo 2 folks that we don't have.

Jim replied with "I'm afraid those two are fairly phylogenetically random...". He said we also find these two as positives in hg E and hg I. I still think it is something to watch if the consistency holds up for both markers across Z253, L1335 and DF49. The two seem to be riding together for L21 purposes so that's something. Is CTS927 in Geno 2?

seferhabahir
12-19-2013, 07:08 PM
Is CTS297 in Geno 2?

I think you mean CTS927 here.

It is not in Geno 2.0 (however CTS297 is, and everyone appears to have an ancestral value of G).

Mikewww
12-19-2013, 08:23 PM
Yes, the CTS927 is the one we thought might be interesting.

Mikewww
12-19-2013, 08:24 PM
The chart at the URL below will eventually replace the "descendancy" tree you are used to seeing. The good news is I've integrated BISDNA's (from Chromo 2) draft work as well as some from FGC to go with what we already know from Geno 2, past testing, etc.

Make sure to look. There are exciting things that Chromo 2 has added to L1065 and M222 in particular.

https://dl.dropboxusercontent.com/u/17907527/R1b-L21_Tree_Chart.jpg

The bad news is I'm going to have to let pretty colors and custom arrangement go. There is just too much and too much more coming to manually position SNPs on a tree anymore. Get use to using the "zoom" button and panning left, right, up and down. I like the way some family tree programs do descendancy trees, but I gave up on creating a GEDCOM format of this. I'm convinced I can programmatically create a GEDCOM file from the table you will see below but that's only worthwhile if free GEDCOM tools can do a nice job of the zooming, panning, unfolding, etc.

The graphic tree chart above should represent the way the R1b-L21_SNPs and R1b-L21_Haplotypes spreadsheets work. Here is the actual table in the SNPs spreadsheet that I'm using to assign haplogroup labels.

https://dl.dropboxusercontent.com/u/17907527/R1b-L21_Tree_SNP_List.pdf

This table is only the "lead with" SNPs. There are many phylogenetic equivalents and identical but differently named synonyms in additional tables down in the SNPindex tab of that same spreadsheet. Some "lead with" SNPs have a long list of equivalents. M222, for example, has over 20.

We could switch to some kind of "R1b1a2a1b. x,y,z" type of long haplogroup labels but even ISOGG is behind so I don't have a long haplogroup label to use in all occasions. Of course, they are behind for good reasons since additional testing is oftentimes needed to confirm accurate placement.

Please note a special thanks should be given to Jim Wilson at BISDNA as I'm using his base "draft" information on some areas of the tree. Keep in mind it's all "draft" and I can and do make errors. Please check it over.

Diverclic
12-19-2013, 09:37 PM
These trees are very much like mtDNA haplogroup trees with very different branch length. The way I take it is to consider the longest branching so far . The other branches might have more SNPs upon sequencing the whole Y (not true with mtDNA).

Mikewww
12-19-2013, 09:44 PM
The chart at the URL below will eventually replace the "descendancy" tree you are used to seeing. ....

https://dl.dropboxusercontent.com/u/17907527/R1b-L21_Tree_Chart.jpg

Converting the powerpoint format to .jpeg seems to cause some much needed loss of resolution. Here's the native powerpoint chart URL.

https://dl.dropboxusercontent.com/u/17907527/R1b-L21_Tree_Chart.pptx

If you don't have Powerpoint, here is MS's free viewer, but I think there are others out there have decent zoom/pan functions.

http://www.microsoft.com/en-us/download/details.aspx?id=13

Mikewww
12-19-2013, 09:48 PM
These trees are very much like mtDNA haplogroup trees with very different branch length. The way I take it is to consider the longest branching so far . The other branches might have more SNPs upon sequencing the whole Y (not true with mtDNA).

We should be careful about looking at a phylogenetic tree like this. There are many SNPs that are phylogenetic equivalents that aren't on the tree. I just pick one "lead with" or defining SNP per branch. There are some 23 SNPs equivalent to M222, according to Wilson at BISDNA. That means that little short hop on the chart from Z2961 through the M222 branch is really quite long. We don't know if M222 is the oldest SNP on that branch or the youngest. It would make quite a difference in this case.

Technically, we should go back to long haplogroup labels with phylogenetic intelligence built in. The tree is really a tree of branches and SNPs are just markers on some of the branches.

Mikewww
12-20-2013, 04:01 AM
The conversion to the .pdf (Adobe Acrobat format) works pretty well. Here it is.

https://dl.dropboxusercontent.com/u/17907527/R1b-L21_Tree_Chart.pdf


Converting the powerpoint format to .jpeg seems to cause some much needed loss of resolution. Here's the native powerpoint chart URL.

https://dl.dropboxusercontent.com/u/17907527/R1b-L21_Tree_Chart.pptx

If you don't have Powerpoint, here is MS's free viewer, but I think there are others out there have decent zoom/pan functions.

http://www.microsoft.com/en-us/download/details.aspx?id=13

Dubhthach
12-20-2013, 09:33 AM
Mike,

I exported your PDF to PNG format, JPEG is a lossy compression format, whereas PNG is lossless. At the same resolution there won't be any blockiness in a png version, pdf is great though for zooming as the fonts scale.

http://compsoc.nuigalway.ie/~dubhthach/R1b-L21_Tree_Chart.png

URL:
http://compsoc.nuigalway.ie/~dubhthach/R1b-L21_Tree_Chart.png

On a Mac you can open a PDF in "Preview" and then select "Export" -> several image formats.

-Paul

hoxgi
12-20-2013, 12:51 PM
Jim replied with "I'm afraid those two are fairly phylogenetically random...". He said we also find these two as positives in hg E and hg I. I still think it is something to watch if the consistency holds up for both markers across Z253, L1335 and DF49. The two seem to be riding together for L21 purposes so that's something. Is CTS927 in Geno 2?

Mike, thanks for checking this out. Jim's reply is somewhat disappointing, but perhaps we will see quite a number of SNPs which are not unique but are still very useful because of their positioning above or below other SNPs which are unique.

You have the list of CTS927+ S2906+ people correct. I wasn't aware of Willis when I posted yesterday.

I have looked for these two SNPs in McClure's FG results; he is Z253+. CTS927 is in his gtype file as ancestral, which doesn't fit with the Chromo 2.0 results, but I'm uncertain how to interpret this result in terms of reliability (number of reads is 1). I cannot find it in either his VariantCompare or HaplogroupCompare files.

S2906 is not listed, but according to the spreadsheet of duplicate names in the Chromo 2.0 results section of the L21 Yahoo group, S2906 is the same SNP as CTS919. CTS919 is also listed in McClure's FG gtype file as negative (number of reads is 0) - but according to FG the ancestral base of this SNP is C, while Chromo 2.0 has S2906 as being ancestral with G and derived with A !

I think all that we can do for now is follow these SNPs as more results come in and see what develops.

Greg H

Mikewww
12-20-2013, 04:49 PM
Mike,

I exported your PDF to PNG format, JPEG is a lossy compression format, whereas PNG is lossless. At the same resolution there won't be any blockiness in a png version, pdf is great though for zooming as the fonts scale.

http://compsoc.nuigalway.ie/~dubhthach/R1b-L21_Tree_Chart.png

URL:
http://compsoc.nuigalway.ie/~dubhthach/R1b-L21_Tree_Chart.png

On a Mac you can open a PDF in "Preview" and then select "Export" -> several image formats.

-Paul

Joss Le G. has asked me to consider positioning the boxes/SNPs by age. It's a great idea and I was actually hoping to use GEDCOM and the birth date as part of the process but all of the GEDCOM viewers I see force a spouse box besides every father only cluttering things up more. I believe I can generate the GEDCOM file format from the SNPindex "tree" but I don't think GEDCOM is the answer anyway. I'll think about Joss' request a bit more. It's good to know .png does ruin the resolution. Thanks. I think the .pdf format is working best, or powerpoint itself.

hoxgi
12-20-2013, 09:21 PM
Neal Downing has confirmed that a member of the Z255 Project who is Z255+ has received his Chromo 2.0 results and is positive for both CTS927 and S2906. Neal and I have been watching for results which may link the Z253 and Z255 clades, as some previous calculations have suggested that they may have separated more recently than the other sons of DF13, and their STR signatures sometimes appear to overlap.

I have asked him to see if this person will upload his results to the L21+ Yahoo Group.

Greg H

Mikewww
12-20-2013, 09:45 PM
Neal Downing has confirmed that a member of the Z255 Project who is Z255+ has received his Chromo 2.0 results and is positive for both CTS927 and S2906. Neal and I have been watching for results which may link the Z253 and Z255 clades, as some previous calculations have suggested that they may have separated more recently than the other sons of DF13, and their STR signatures sometimes appear to overlap.

I have asked him to see if this person will upload his results to the L21+ Yahoo Group.

Greg H

The first run through (about the only) that I ran with 67 STR data sets on Ken Nordtvedt's interclade TMRCA estimator put Z255 and Z253 as closer related to each other than to other DF13 major subclades. I drew up a tentative tree that linked these two on their own branch before rooting into DF13.

Here it is from almost two years ago. I wouldn't have ever thought DF49 would be tied in though. At least from DF23 on down, its distinctive.
https://dl.dropboxusercontent.com/u/17907527/R1b-L21_Tree-Timeline.pptx

You may really be on to something. Jim Wilson seems a little reticent on this but two SNPs consistently sticking together could easily be a good indicator. Statisticians could probably figure out how two unstable SNPs compare to one "thought to be" stable SNP. I think this is the future with SNPs and more SNPs. It will more important to look at an SNP in context of its placement on a potential tree visa vi others.

Dubhthach
12-21-2013, 08:30 AM
Rich Stevens has confirmed to me via email that he is negative for these two SNP's. This would rule out DF41 as been of this "Super block" under DF13. Interesting times!

Do we know the status of L371, CTS4466 and Z251? Let alone the smaller snp's under DF13.

-Paul
(DF41+)

corner
12-21-2013, 12:13 PM
Mike, thank you for the SNPs spreadsheet, which is very helpful already in analysing new SNPs.

Could you please check it for two SNPs from Chromo 2.0, namely CTS927 and S2906? Of the 12 current Chromo 2.0 results in the L21 Yahoo group, CTS927 and S2906 are positive in the two Z253+ kits, the three L1335+ kits and the one DF49+ kit, whereas your spreadsheet only has four positive results for each when I use the filter.

These two SNPs are negative in the other Chromo 2.0 results (one DF13*, three L513+, one DF21+ and one DF41+) - except for a no-call on CTS927 in the DF21+ kit.

So they appear to define a branch of DF13 which then divides into Z253, L1335 and DF49.

Joss Le Gall's results are inconsistent with the above, as he is Z253+ and negative for these two SNPs, but only his initial results are available at present, not the updated ones.

Greg H

I am positive for the SNPs you mention, CTS927 and S2906, but L21 negative. I match the DF27** Rox2 cluster and just received Chromo2 Raw results, still going through them. Another SNP that stands out (maybe a connection?) is an incongruous positive result for a SNP that looks usually to be found downstream of L21 (under Z253) named PF825. Named as, 'R1b-L430 (R1b-PF825)' in the Morley Subclade Predictor.

hoxgi
12-21-2013, 12:14 PM
Do we know the status of L371, CTS4466 and Z251? Let alone the smaller snp's under DF13.

-Paul
(DF41+)

I have only looked at the Chromo 2.0 results available in the L21 Yahoo Group files; I don't have access to any other Chromo 2.0 results. So I haven't seen any results from the smaller subclades you have listed.

CTS927 and S2906 are both negative in one DF13*, three L513+, one DF21+ and one DF41+ kit (Rich Stevens), with the exception of a no-call on CTS927 in the DF21+ kit. Clearly we need substantially more results to see if these two SNPs continue to give reliable results across the DF13 subclades.

Ubfortunately I have a technical problem with S2906. According to both ISOGG and Britain's DNA, S2906 is the same SNP as CTS919. ISOGG identifies both by the Ch37 reference 7087086, with the mutation being C>T, and this is consistent with FG results for CTS919. However Chromo 2.0 gives S2906 results as GG ancestral, AG derived, so presumably G>A, which is inconsistent with the above sources.

Greg H

Mikewww
12-27-2013, 04:00 AM
Here is what I'm using as working draft trees for R1b-L21. This first format is an excerpt from the larger spreadsheets. It contains the actual rules that govern how the spreadsheet assigns haplogroup labels. It is in a list/table format.

https://dl.dropboxusercontent.com/u/17907527/R1b-L21_Tree_SNP_List.pdf

This graphic is simply a descendancy tree version of the list/table. There is a chance I could have errors in it as I manually build it from the list/table format above.

https://dl.dropboxusercontent.com/u/17907527/R1b-L21_Tree_Chart.pdf

Well, anything could have errors in it.... Let me know and I'll try to keep up with corrections and updates.

You can also find these under Links, under More from the R1b-L21-Project yahoo group.

The following were part of the input I used. They are draft trees that Dr. James F. Wilson of BISDNA has created. He is working on additional subclades of L21/S145. Chromo 2 is adding a lot of sorely needed resolution under M222.

https://dl.dropboxusercontent.com/u/17907527/R1b-L21_Tree_Chart_M222_JFW.pptx

https://dl.dropboxusercontent.com/u/17907527/R1b-L21_Tree_Chart_S530_L1335_JFW.pptx

https://dl.dropboxusercontent.com/u/17907527/https://dl.dropboxusercontent.com/u/17907527/R1b-L21_Tree_Chart_S1136_CTS4466_JFW.pptx

wlharris1055@yahoo.com
12-27-2013, 11:06 AM
Here is what I'm using as working draft trees for R1b-L21. This first format is an excerpt from the larger spreadsheets. It contains the actual rules that govern how the spreadsheet assigns haplogroup labels. It is in a list/table format.

https://dl.dropboxusercontent.com/u/17907527/R1b-L21_Tree_SNP_List.pdf

This graphic is simply a descendancy tree version of the list/table. There is a chance I could have errors in it as I manually build it from the list/table format above.

https://dl.dropboxusercontent.com/u/17907527/R1b-L21_Tree_Chart.pdf

Well, anything could have errors in it.... Let me know and I'll try to keep up with corrections and updates.

You can also find these under Links, under More from the R1b-L21-Project yahoo group.

The following were part of the input I used. They are draft trees that Dr. James F. Wilson of BISDNA has created. He is working on additional subclades of L21/S145. Chromo 2 is adding a lot of sorely needed resolution under M222.

https://dl.dropboxusercontent.com/u/17907527/R1b-L21_Tree_Chart_M222_JFW.pptx

https://dl.dropboxusercontent.com/u/17907527/R1b-L21_Tree_Chart_S530_L1335_JFW.pptx

https://dl.dropboxusercontent.com/u/17907527/https://dl.dropboxusercontent.com/u/17907527/R1b-L21_Tree_Chart_S1136_CTS4466_JFW.pptx

Hello Mike. I noticed your new R1b-L21 Descendant Tree has CTS3655 as a main branch off DF21 whereas the DF21 Project Tree Chart of David R shows it under Z246. My SNP options are limited to L130 and L641/L642 under David' Chart; but, would include CTS 3655 under your chart. Which is correct?

Thanks and ......... Merry Christmas!
Bill

Mikewww
12-27-2013, 05:12 PM
Hello Mike. I noticed your new R1b-L21 Descendant Tree has CTS3655 as a main branch off DF21 whereas the DF21 Project Tree Chart of David R shows it under Z246. My SNP options are limited to L130 and L641/L642 under David' Chart; but, would include CTS 3655 under your chart. Which is correct?

Thanks and ......... Merry Christmas!
Bill

I think mine is incorrect. I notice I didn't even have DF5 on there. The SNPindex list is really the one I'm using. The graphic chart is just an effort to communicate it in a more recognizable format. I'm going to look at DF21's branching in a few minutes.

wlharris1055@yahoo.com
12-27-2013, 07:36 PM
I think mine is incorrect. I notice I didn't even have DF5 on there. The SNPindex list is really the one I'm using. The graphic chart is just an effort to communicate it in a more recognizable format. I'm going to look at DF21's branching in a few minutes.

Yeah well my mistake too .... I meant my SNP options are limited to L130 and S424 and S426 on David' Chart .... not L641/L642.

Mikewww
12-27-2013, 08:00 PM
Yeah well my mistake too .... I meant my SNP options are limited to L130 and S424 and S426 on David' Chart .... not L641/L642.

I think I've got it fixed now.
https://dl.dropboxusercontent.com/u/17907527/R1b-L21_Tree_Chart.pdf

Mikewww
01-24-2014, 05:11 PM
There are some minor updates under M222 now.
https://dl.dropboxusercontent.com/u/17907527/R1b-L21_Tree_Chart.pdf

Mikewww
02-24-2014, 11:43 PM
The biggest news lately is that FGC5496 now has several layers of branches. This is going to start getting wild soon!
https://dl.dropboxusercontent.com/u/17907527/R1b-L21_Tree_Chart.pdf[/QUOTE]

Mikewww
02-24-2014, 11:47 PM
Subtopic: Submitting candidate SNPs for placement on ISOGG tree

I don't think I can boil the "all haplogroups from Adam" ocean, but hopefully I can help L21 folks be consistent, efficient and well documented.

I've emailed individuals about this and posted this topic at ISOGG and a couple of other forums. I haven't received much of a response so I'll go ahead and promote this as a proposal. This will become very important over the next several months as more SNPs are identified and positioned. We need need a consistent and complete process for submission with full disclosure. The consistency and completeness reduce errors and improve efficiency. Full disclosure is needed so that everyone can review and repeat/re-assess the data and get the same results.

I propose that all email requests to ISOGG for SNPs to added to the ISOGG R1b-L21 tree use the format below and carbon copy the L21 yahoo group at R1b-L21-Project@yahoogroups.com

Along with this I'll try to start being consistent on terminology. When I use the term "candidate" SNP I'm referring to the SNP that is being submitted as formal subclade/haplogroup marker on the Y DNA tree. I use the term "peer" to refer to a brother(s) (not literally, though) SNPs already on the formal Y DNA tree. A peer's lineage is parallel and exclusive to the candidate SNP but both the candidate and the peer have the same "immediate ancestor" on the Y tree. The "immediate ancestor" could be viewed a the father or parent.

By copying this L21 yahoo group, we will have a log that will automatically be archived and retrievable.

I'm not in love with any of this process or terminology so if someone has a better way, let's discuss this, but now is the time; before the avalanche, flood, tsunami or whatever you want to call it hits.

----------- Subject line of email --------------------------
SNP xyz (candidate SNP) Submission for ISOGG tree


----------- Outline format of body of email -------------------------


Candidate SNP and synonym labels
----------------------------------------------
SNP labels, URL to physical address info

Candidate SNP diversity status
------------------------------------------
Kit #, Surname, Project URL, candidate SNP derived result, Genetic Distance from comparison
Kit #, Surname, Project URL, candidate SNP derived result (for comparison target)

Immediate Ancestor (father) SNP currently on ISOGG chart
------------------------------------------------------------------
SNP

Ancestral and derived haplotypes proving descendant status
------------------------------------------------------------------------------
Kit #, Surname, Project URL, ancestor SNP derived result, candidate SNP deriived result
Kit #, Surname, Project URL, ancestor SNP derived result, candidate SNP ancestral result

Peer (brother) SNP currently on ISOGG chart (1 through n) (if applicable)
-----------------------------------------------------------------
Kit #, Surname, Project URL, peer SNP derived result, candidate SNP ancestral result

Descendant (child) SNP currently on ISOGG chart (1 through n) (if applicable)
-----------------------------------------------------------------
Kit #, Surname, Project URL, descendant SNP derived result, candidate SNP derived result
Kit #, Surname, Project URL, descendant SNP ancestral result, candidate SNP derived result

razyn
02-25-2014, 12:13 AM
Re:
Immediate Ancestor (father) SNP currently on ISOGG chart I have a bad feeling that that could get to be like older DAR memberships, based on the genealogies submitted by more or less closely related members in the 1880s... etc. Same phrase is used for Peer and Descendant. It's OK, if the current ISOGG chart is perfect. Otherwise it could become problematic.

hoxgi
02-25-2014, 12:04 PM
Subtopic: Submitting candidate SNPs for placement on ISOGG tree

It may be worth including the tests used for each SNP, both positive and negative, in the ISOGG submission. Typically, ISOGG has required both positive and negative SNP tests to be confirmed by Sanger sequencing, which can generate considerable extra expense. However, perhaps results obtained by other methods may now be considered on a case by case basis.

For example, CTS9881, which is positive in the Irish Type 4 subclade of L1066, has recently been added to the ISOGG Y-tree. As it is a new terminal SNP, it was necessary to demonstrate that it is downstream from L1066 and satisfies the GD requirement for ISOGG listing. We had two positive results, one from Geno 2.0 and one from Chromo 2, separated by an appropriate GD. Only one of these has been confirmed by Sanger sequencing. There were multiple negative results in L1066+ persons who were not Irish Type 4, obtained from both Geno 2.0 and Chromo 2 results, but not Sanger testing.

ISOGG accepted these results, probably because of several factors, including the concordant results between Geno 2.0 and Chromo 2, the large number of negative results (at least twelve) in L1066+ persons (which effectively eliminated any concern regarding no-calls or technical errors), and the correlation between the positive results and an already well-defined (by STRs) grouping. No further Sanger testing was required.

As there will probably be a large number of new SNPs requiring submission to ISOGG, it may be worth considering what Sanger confirmation is required in each case, and if necessary requesting clarification from ISOGG. I think Mark Jost, the ISOGG co-ordinator for L21, is going to become very busy.

Greg H

RobertCasey
02-25-2014, 03:41 PM
Subtopic: Submitting candidate SNPs for placement on ISOGG tree


----------- Subject line of email --------------------------
SNP xyz (candidate SNP) Submission for ISOGG tree


----------- Outline format of body of email -------------------------


Candidate SNP and synonym labels
----------------------------------------------
SNP labels, URL to physical address info

Candidate SNP diversity status
------------------------------------------
Kit #, Surname, Project URL, candidate SNP derived result, Genetic Distance from comparison
Kit #, Surname, Project URL, candidate SNP derived result (for comparison target)

Immediate Ancestor (father) SNP currently on ISOGG chart
------------------------------------------------------------------
SNP

Ancestral and derived haplotypes proving descendant status
------------------------------------------------------------------------------
Kit #, Surname, Project URL, ancestor SNP derived result, candidate SNP deriived result
Kit #, Surname, Project URL, ancestor SNP derived result, candidate SNP ancestral result

Peer (brother) SNP currently on ISOGG chart (1 through n) (if applicable)
-----------------------------------------------------------------
Kit #, Surname, Project URL, peer SNP derived result, candidate SNP ancestral result

Descendant (child) SNP currently on ISOGG chart (1 through n) (if applicable)
-----------------------------------------------------------------
Kit #, Surname, Project URL, descendant SNP derived result, candidate SNP derived result
Kit #, Surname, Project URL, descendant SNP ancestral result, candidate SNP derived result

Here is a table of a summary of ISOGG qualification that I have used at my L21 web site. With so many duplicate
YSNPs with Full YChr tests, we need to post those YSNPs as well.





SNP
Number
FTDNA Submission Number
(or WTY GRC number)
Project Name
Test Results



Prove ISOGG Father of DF73





DF73+

58625

L21 Plus


Z2534+




Prove DF73 is not a duplicate of Z2534





DF73-

77349

L21 Plus


Z2534+




Eliminate possible fathers of DF73





DF73+

58625

L21 Plus


L226-




DF73+

58625

L21 Plus


L643-




DF73+

58625

L21 Plus


Z2185-




Eliminate possible sons of DF73





DF73-

77349

L21 Plus


L226+




DF73-

57521

L21 Plus


L643+




DF73-

231060

L21 Nat Geo


Z2185+




Prove duplicate SNPs of DF73





NA

None known

NA


NA




Prove SNPs that are sons of DF73





NA

None known

NA


NA




Prove 10 % Diversity (ISOGG)





DF73+


58625

L21 Plus


Base




DF73+


143916

L21 Plus


GD=16





Of course, a Project URL may not be a FTDNA YSNP report. It could be Big Y file, FG file, Nat Geo file, Walk the Y file, CROMO2, YSEQ, etc. I am not
sure that Surname will ever be required but it does enhance the information as YSNPs get genealogical - yet still make the GD qualification. I think
that duplicate YSNPs need to be separated from private sons - this is a major improvement in the ISOGG haplotree for tracking proven private sons
that do not yet qualify for the diversity requirement vs. just lumping them into duplicates. This would be much better improvement than surnames.

Mikewww
02-25-2014, 04:45 PM
....
Of course, a Project URL may not be a FTDNA YSNP report. It could be Big Y file, FG file, Nat Geo file, Walk the Y file, CROMO2, YSEQ, etc.
Agreed. I just think there needs to be some link to a public testing institution web site so the authority standing behind the test result is confirmed.


.... I am not sure that Surname will ever be required but it does enhance the information as YSNPs get genealogical - yet still make the GD qualification.
I don't think it is required. I added it as a checkpoint with the kit#/ID#. I've found that is always good to include a name with the ID as occasionally digits get transposed.


. I think
that duplicate YSNPs need to be separated from private sons - this is a major improvement in the ISOGG haplotree for tracking proven private sons
that do not yet qualify for the diversity requirement vs. just lumping them into duplicates. This would be much better improvement than surnames.
I'm not sure I'm following you completely on this.

The format I suggested does not really support phylogenetically equivalent SNPs. I just want to get something on board for positioning/insertion of newly discovered SNPs. Equivalent SNPs are another matter that is one of gray areas. SNPs are artificially deemed equivalent at some point but I don't know of a minimum testing requirement.

RobertCasey
02-25-2014, 05:32 PM
I'm not sure I'm following you completely on this.

The format I suggested does not really support phylogenetically equivalent SNPs. I just want to get something on board for positioning/insertion of newly discovered SNPs. Equivalent SNPs are another matter that is one of gray areas. SNPs are artificially deemed equivalent at some point but I don't know of a minimum testing requirement.

I agree that a URL link is key (citing source references is always desirable). I agree that adding surname would help - but this is unlikely to fly with ISOGG.

As far as equivalent YSNPs, ISOGG does accept these and posts hundreds of them. We should continue to submit known equivalent YSNPs. I am thinking about adding Z252 as an equivalent of Z253. I think this adds value to the haplotree for older equivalents so people do not continue discover it is an equivalent. There are around ten 1K tests and two FG tests that indicate that Z252 is equivalent of Z253 as well as many other L21 tests that indicate Z252 is not between Z253 and L21. Maybe there should be a separate form for older equivalents. Sorting the equivalents out is a major task that should be documented - at least for older YSNPs that have descendants. ISOGG is the best place to track all the different names of the same mutation as well as all those equivalents as well.

As far as "private" YSNPs under terminal YSNPs, I think ISOGG needs to at least track them in some manner - but like surnames, that is unlikely. I think this would a enormous improvement to add a list equivalents (similar to how Thomas Krahn listed many YSNPs as approximate to YSNPs). They may proven as sons or ancestors to the approximate YSNP later - but the ISOGG haplotree could provide a great place to put those thousands of YSNPs that are under research.

Mikewww
02-25-2014, 07:02 PM
I agree that a URL link is key (citing source references is always desirable). I agree that adding surname would help - but this is unlikely to fly with ISOGG..
I'm not advocating adding surnames to the ISOGG trees or anything like that... although that would be interesting to think about in the future with much better phylo tree mapping technologies (hopefully).

I'm only asking ISOGG for guidance on how they want to see submissions for branch creating SNP insertions into tree.



As far as equivalent YSNPs, ISOGG does accept these and posts hundreds of them. We should continue to submit known equivalent YSNPs. I am thinking about adding Z252 as an equivalent of Z253. I think this adds value to the haplotree for older equivalents so people do not continue discover it is an equivalent. There are around ten 1K tests and two FG tests that indicate that Z252 is equivalent of Z253 as well as many other L21 tests that indicate Z252 is not between Z253 and L21. Maybe there should be a separate form for older equivalents. Sorting the equivalents out is a major task that should be documented - at least for older YSNPs that have descendants. ISOGG is the best place to track all the different names of the same mutation as well as all those equivalents as well. I agree that equivalent SNPs must be considered and need to be added to the tree. I just didn't have that as a first priority vis a vis SNP insertion/branch creation. I'm a little unclear on the definition of when equivalency can be assumed. Obviously duplicate testing (duplicate of the original branch marking SNP) across peers and children is required depending on the situation. However how many Lead with SNP + Equivalent SNP + and - - results are needed to assume equivalency is something I don't know. Making the equivalency assumption is a game of probabilities but can only be truly understood in hindsight.


As far as "private" YSNPs under terminal YSNPs, I think ISOGG needs to at least track them in some manner - but like surnames, that is unlikely. I think this would a enormous improvement to add a list equivalents (similar to how Thomas Krahn listed many YSNPs as approximate to YSNPs). They may proven as sons or ancestors to the approximate YSNP later - but the ISOGG haplotree could provide a great place to put those thousands of YSNPs that are under research. Private SNPs may be the most important of all and it is mind boggling to manage in a broad way. I think every youthful subclade needs its own advocate.

L21xDF13xDF63
DF63
DF13xBig10
DF49xM222
M222
DF21
DF41
L513
L1335
CTS4466
Z251
Z253
Z255
FGC5496
... more to come.

BTW, I think Mark J is our L21 ISOGG coordinator. May God bless him!

Mikewww
02-25-2014, 07:27 PM
It may be worth including the tests used for each SNP, both positive and negative, in the ISOGG submission. Typically, ISOGG has required both positive and negative SNP tests to be confirmed by Sanger sequencing, which can generate considerable extra expense. However, perhaps results obtained by other methods may now be considered on a case by case basis.

For example, CTS9881, which is positive in the Irish Type 4 subclade of L1066, has recently been added to the ISOGG Y-tree. As it is a new terminal SNP, it was necessary to demonstrate that it is downstream from L1066 and satisfies the GD requirement for ISOGG listing. We had two positive results, one from Geno 2.0 and one from Chromo 2, separated by an appropriate GD. Only one of these has been confirmed by Sanger sequencing. There were multiple negative results in L1066+ persons who were not Irish Type 4, obtained from both Geno 2.0 and Chromo 2 results, but not Sanger testing.

ISOGG accepted these results, probably because of several factors, including the concordant results between Geno 2.0 and Chromo 2, the large number of negative results (at least twelve) in L1066+ persons (which effectively eliminated any concern regarding no-calls or technical errors), and the correlation between the positive results and an already well-defined (by STRs) grouping. No further Sanger testing was required.

As there will probably be a large number of new SNPs requiring submission to ISOGG, it may be worth considering what Sanger confirmation is required in each case, and if necessary requesting clarification from ISOGG. I think Mark Jost, the ISOGG co-ordinator for L21, is going to become very busy.

Greg H

We have seen that the Geno2/Chromo2 technology level produced some strange results and in some cases many no calls. Supposedly, Big Y (and I would think FGC hi-quality results) are supposed to have a more accurate and reliable testing. I don't know if that is true, but if it is that could change things.

We've seen on other threads that Warwick has proposed that it may not be economic to run one at a time Sanger Sequencing for all candidate SNPs.

I don't know, but I think the conservative approach, which Mark J has advocated, is the Sanger Sequencing validation. That would make Thomas Krahn/YSEQ happy. I've got my sample "in the mail" to them with hopes of having some stuff to "make a wish" for and have one at a time testing options. I have our L21 data (STRs and SNPs) integrated through YSEQ's ID=11 R1b-L21 public display.

razyn
02-25-2014, 07:30 PM
BTW, I think Mark J is our L21 ISOGG coordinator. May God bless him!

ISOGG Notes section says
Changed R1b-P312 contact person to Mark Jost on 6 February 2014.

Not just L21, that is.

Mikewww
02-26-2014, 01:38 AM
ISOGG Notes section says "P312"

Not just L21, that is.

He needs a double or triple dose of favor then.

Mikewww
03-27-2014, 09:57 PM
Just an update. I'm NOT updating the R1b-L21_Tree_Chart powerpoint, .png and .pdf files for the time being. There is just too much change to try to worry about drawing it up manually as things shift. The new Chromo 2 anonymized file has been very helpful but we don't know have a clear view into the identities and haplotypes as well as how those SNPs (phylogenetically) relate to new FGC and Big Y identified SNPs. I'll update the graphical trees, but not for a while.

Several people are doing great work on analyzing information. I am convinced this needs to be done at lower and lower (more youthful) levels of the tree taking us down to DF13's big ten subclades, DF13* and DF63's subsclades and then even more youthful into the genetic genealogy realm.

I think there is a huge point of leverage with the 340 plus Big Y results we will see for L21. This gives us a good and consistent scan of the Y chromosome (what FTDNA calls the "gold" regions). What's important is that sets up for a nice apples to apples comparison which is what is needed for phylogenetic tree identification (comparative methods - click here. (http://en.wikipedia.org/wiki/Phylogenetic_comparative_methods))

That being said, I'm working on what I call R1b-L21_SNP_Tree_Discovery. This file is a quick view of all 104 (so far but will go up to 350) L21 derived variants from the "gold" regions. This file can be access from the R1b-L21-project yahoo group. There is a have a classification system for SNPs (i.e. single individual, single surname, multi-family, public consistent, public unsure, etc.) I include STR signature variety labels for individuals. In some cases the variety labels are being destroyed, but usually they are validated. Eventually, people at 111 STRs should be able to get a good idea of what SNPs they want to test for if they want to do a one at a time (a la carte) approach.

After I'm done with that, I'll review all derived simple SNPs (straight allele one for one changes) by individual in a comparison. This won't be apples to apples but sets up for inclusion of comparisons with other test packages, including reverse engineering one at a time SNP results from FTDNA and YSEQ or Geno 2 and Chromo 2 where people are sharing their data.

Anyway, it goes without saying. If you are L21+, join the R L21 project. Join other pertinent projects but the L21 project is a very large database in and of itself so that is the best place (IMO) for up to date info. Plus... Richard S is the fastest haplogroup project administrator around.

Mikewww
03-28-2014, 06:40 PM
I'll try to cross-post here (from the L21 yahoo group) when get breaks.

Here is an update - Are DF41 and Z255 brothers under super-subclade of DF13?

This variant is not named but it is consistent from a phylogenetic standpoint so far (everything is always "so far".)
16132308T>TGATA

It appears in all 7 Z255 guys and 3 of 6 DF41 guys. The other 95 L21 Big Y folks don't have it reported as derived.

I checked the .vcf files on the other 3 DF41 guys that don't have it. That location is not in those file so we don't know if they have it or don't. It's possible then, that 16132308T>TGATA indicates DF41 and Z255 are brothers.

Z255 is not that old and it is related to L159.2 (not sure exactly how, though). The Irish Sea/Leinster/Lagin STR varieties people are the biggest part of it.

DF41 has a few different signatures some of which may be Welsh and some of which may be Scottish and other Breton. I'm sure some are Irish too and something else in the British Isles.

Williamson
03-28-2014, 07:07 PM
Here is an update - Are DF41 and Z255 brothers under super-subclade of DF13?

This variant is not named but it is consistent from a phylogenetic standpoint so far (everything is always "so far".)
16132308T>TGATA

It appears in all 7 Z255 guys and 3 of 6 DF41 guys. The other 95 L21 Big Y folks don't have it reported as derived.

16132308 is an STR. In the reference sequence, there are 7 repeats of GATA at that location. It may be a fairly stable STR, but I don't think it represents any common connection here. Rory Cain also has 8 repeats of GATA at that location, and he's R-DF21. Irwin (75606) in CTS4466 has only 6 repeats. He has the opposite mutation, a deletion rather than an insertion.

Alex

Mikewww
03-28-2014, 07:27 PM
16132308 is an STR. In the reference sequence, there are 7 repeats of GATA at that location. It may be a fairly stable STR, but I don't think it represents any common connection here. Rory Cain also has 8 repeats of GATA at that location, and he's R-DF21. Irwin (75606) in CTS4466 has only 6 repeats. He has the opposite mutation, a deletion rather than an insertion.

Alex
I understand it is not a straight and simple SNP and therefore should be less reliable. That's why I switched to using the same terms as the NGS testing offerings use - variants.

However, its very consistent so far within L21, at least on the Big Y reporting. It might be a phylogenetic indicator, yet.

If it is a parallel mutation with DF41, it might be useful for the Stewart folks, where I noticed it.

R.Rocca
04-08-2014, 07:59 PM
I'll try to cross-post here (from the L21 yahoo group) when get breaks.

Here is an update - Are DF41 and Z255 brothers under super-subclade of DF13?

This variant is not named but it is consistent from a phylogenetic standpoint so far (everything is always "so far".)
16132308T>TGATA

It appears in all 7 Z255 guys and 3 of 6 DF41 guys. The other 95 L21 Big Y folks don't have it reported as derived.

I checked the .vcf files on the other 3 DF41 guys that don't have it. That location is not in those file so we don't know if they have it or don't. It's possible then, that 16132308T>TGATA indicates DF41 and Z255 are brothers.

Z255 is not that old and it is related to L159.2 (not sure exactly how, though). The Irish Sea/Leinster/Lagin STR varieties people are the biggest part of it.

DF41 has a few different signatures some of which may be Welsh and some of which may be Scottish and other Breton. I'm sure some are Irish too and something else in the British Isles.

Just an FYI...16132308 looks like it may be an STR and I have found the same value (T>TGATA) in a U152+L2+ sample.

Mikewww
05-13-2014, 10:30 PM
I haven't tried to update the R1b-L21_Tree_Chart for a while. It was hopelessly behind as new NGS results came in weekly from Big Y. I still don't have a Big Y result, myself. There was a sample issue, but gee whiz, its about time. We've hit the 300 mark of Big Y results that I can get my hands on.

I've tried to use the available data and analysis, and assessments from folks like Williamson, Jost, McCown, Brown, Wright, Hockings, Krahn and others that I'm forgetting but here is the best I can do right now.

https://dl.dropboxusercontent.com/u/17907527/R1b-L21_Tree_Chart.png

The amazing thing is we are finding zero equivalents for DF13 and we are now at 21 direct branches from DF13 and counting. DF13 seems to be providing the definition of a bush, albeit from a single trunk, but no scraggily tree that's for sure. The DF13 TMRCA's family must have been active at something.

R.Rocca
05-20-2014, 03:06 AM
I haven't tried to update the R1b-L21_Tree_Chart for a while. It was hopelessly behind as new NGS results came in weekly from Big Y. I still don't have a Big Y result, myself. There was a sample issue, but gee whiz, its about time. We've hit the 300 mark of Big Y results that I can get my hands on.

I've tried to use the available data and analysis, and assessments from folks like Williamson, Jost, McCown, Brown, Wright, Hockings, Krahn and others that I'm forgetting but here is the best I can do right now.

https://dl.dropboxusercontent.com/u/17907527/R1b-L21_Tree_Chart.png

The amazing thing is we are finding zero equivalents for DF13 and we are now at 21 direct branches from DF13 and counting. DF13 seems to be providing the definition of a bush, albeit from a single trunk, but no scraggily tree that's for sure. The DF13 TMRCA's family must have been active at something.

Thanks Mike. Are these only Sanger sequenced SNPs or perhaps only the ones that have been given SNP names?

Mikewww
05-21-2014, 12:12 AM
Thanks Mike. Are these only Sanger sequenced SNPs or perhaps only the ones that have been given SNP names?

Many are but this seems to be changing daily as Thomas looks through these and evaluates them. Generally, the ones that are named were picked out of blocks (equivalents) of SNPs for their expected reliability in testing. There are more potential branches identified. The ones on the tree appear to be consistent and reliable, but there are many equivalents behind them. I should note that the exceptions are labeled in red font. In one case, BISDNA thinks the SNP is consistent and YSEQ does not like the reliability factors surrounding it and we don't have the equivalents to use as tie-breakers.

We are up to 209 distinct branches. We've got close to another 20 Big Y's coming in today so I expect to have a little more movement. An example is in my own cluster where we have a block of 15 or so potential equivalents shared by two people. A third individual (Swedish no less) is coming in today so I'd be surprised if he doesn't break up this block in some way. The key is multiple people in a cluster and neighboring clusters need to test. This cuts to the chase.

My graphic is intended to be a "simple" descendants tree like view.
http://www.littlescottishcluster.com/RL21/NGS/Tree.html

However, to get an idea of the phenomena of blocking of equivalent SNPs, it might be helpful to look at Alex Williamson's big tree perspective. It's harder to visualize and its a bit behind but it is actually much more informative. You have to go to the bottom to find the scroll bar to move right and left. It's a very wide chart.
http://www.littlescottishcluster.com/RL21/NGS/Tree.html

The strings or blocks of SNPs provide some relative time value, apparently the result of lineage bottlenecks.

I tried to do this on a graphic tree for L21 but there is no amount of zooming/panning that seems to work well on the tiny fonts required.

Here's my cut at what I'm calling an SNP "depth" chart for just L513. Only shared SNPs are shown, but for mutation rate guys, FYI, there seem to be an average of 37 of the Big Y gold/passed SNPs per lineage since the L513,CTS5396,Z249,S5191,S1594 block that marks the L513 subclade.
https://dl.dropboxusercontent.com/u/17907527/R1b-L513_Tree_SNP_Chart.pdf

WOLFF éric
07-26-2015, 12:38 PM
What´s with R1b-DF13*-Y14049 any informations now ?