24 February 2015 - Z29424 update
Since Ray identified Z29424 as a tentative new subgroup based on Mr Del and Mr Herbet's results, we have now twice been able to expand its potential coverage. First, Mr Beyeler's Full Genomes results showed he shares this mutation, and today's results for Mr Hopkins indicate the same. In Mr Hopkin's case, this means that we can extrapolate to all men from the STR-defined DYS534<15 subgroup. With four BigY/FG results in and one more BigY on order we should be able to improve our knowledge of this branch in the near future.
10 February 2015 - Update from Ray Banks (via the Haplogroup G Facebook page)
UPDATE ON A U1 SUBGROUP
I'm confident that L654.2 will be the first confirmed subgroup of L1264, based on Mr. Nochevnoy's Big Y results. Mr. Wuest whose Walk through the Y results do not display publicly shares this mutation. I thought there was another man with confirmed L654.2 and started doing all the paperwork for L654 (not so), and I will leave these partially done as we test Mr. Nochevnoy for confirmation. The two men with this mutation are not closely related so it may have wider coverage.
28 January 2015 - Update from Ray Banks (via the Haplogroup G Facebook page)
I added a tentative new subgroup to the U1 area of the G tree tonight, Z29424. It joins the Z2017 and L1263 subgroups under Z2003. The two men with Z29424 are not closely related so it may be a big subgroup. One has Italian and the other Turkish Cypriot ancestry. It will take multiple months to get this validated.
1 January 2015 - 2014 overview from Ray Banks (via the Haplogroup G Facebook page)
YEAR END SUMMARY - Haplogroup G
I will shortly post a summary on various topics. A happy New Year to everyone who uses 1 Jan as the start of the year. I want to thank all the project administrators who have done wonderful work this year, and the project members who have pitched in to learn more about their male ancestors. A special thanks also to all to the donors who have made it possible to validate new G subgroups which tell the story of our ancestors’ migrations. It is costing about $400 a month to sponsor the testing, and donations of $5 or more are most welcome to the G Group Fund at YSeq.http://shop.yseq.net/group_alleles.php?gid=48
YEAR END SUMMARY – Research Summary
We added abt 25 new subgroups to the ISOGG G Tree this year (a few still not added), about one every other week. But the rate of additions is increasing near year end after long testing is finally nearing completion. I also added perhaps about 1000 equivalents SNPs to the tree based on next generation sequencing.
This year I also upgraded the experimental G tree so that it is now better organized and more of a Composite G tree that shows all the subgroups suggested in other sources. There are about 200 G subgroups listed
I have the only analysis available where all samples from all sources can be analyzed simultaneously.
Because of the work on the overall Composite Y Tree since September and a completion expected only next month, there is some backlog in Big Y analyses. I should catch up when the initial Composite Tree is completed. I also cover haplogroups C, D and E1a, and their files have many more mutations and take longer to analyze.
The Big Y tests ordered seem mostly to be ordered by men who do not have a good idea of what the testing is about. And they are not always the ideal next person needed for improving our knowledge of G subgroups. But we have, nevertheless, been able to come up with some quite useful for a strong majority of the samples
YEAR END SUMMARY – The DNA companies
Family Tree DNA unveiled a new tree this year. This has not been a great success because it was based overwhelmingly on Geno 2.0 testing, and some of these new tree subgroups are false positives. They have had to revise the tree multiple times to eliminate items which in subsequent testing did not pan out. In the administrator’s listings, incorrect category names persist. Dr. David Mittelman indicated early in the year that the tree would be revised by the end of 2014 to include items not listed in Geno 2.0. This has not happened.
The Big Y has turned out to be a do-it-yourself analyses test because their analyses is not very useful because of the inclusion of lots of useless items reported as new to the customer and used in showing “matches” who often just have the most useless mutations. There is no info as to whether they will try to aggregate the data to show a tree.
Perhaps it is with this in mind that they have announced a panel of SNPs soon to be available. I provided the best listing I could by way of suggestions that would make the test very useful, but it is unclear how many of these will make it into this panel of tests, called a deep clade test.
Several of the administrators made pushes toward the end of the year to have members order individual SNP tests for $39. Rolf Langland’s L497 Project was a stand out.
Full Genomes Corp. This company made available several options for sequencing, one being less comprehensive and nearer to the Big Y price. They have reported hundreds of samples being processed. Our G men should be abt 5% of the total, but we have not gotten any G results this year. Correspondents have mentioned 5 samples being tested there, but no results yet. Full Genomes also began analyzing Big Y data files for $50.
YFull This company also now provides Big Y analyses for $50. The services of both these companies for analysis can be particularly useful where only a few SNPs were found. Family Tree does not report valid SNP mutations when Big Y produces less than 11 reads out of the typical 50.
All of the analyses seem to have a defect in their software in which an item will be reported as positive or negative when the majority of the reads show that. But I have found multiple items in such situations where the customer has have individual testing of the site with a different result. This is not common, but checking the BAM file (raw results) is important to identify this situation.
ySeq We have tested at least 100 new candidate SNPs there with good results. They also offer testing of these to the general public. There are about 240 G items offered, but several dozen are items requested by 2 men which have never been tested and do not have general utility.
DNATribes is a new company that include Y testing as part of its test. It purports to use a different method of testing, but we have not seen results yet
12 December 2014 - BigY update
As per Ray's update below, we appreciate your patience once your BigY is completed. Analysing the results is very time consuming if done properly, and Ray is personally handling a lot of them in this manner...
On a very positive note, we now have 9 BigY's either delivered or on order (6 in L13, 3 in L1266) so we are in a good position to advance our research in more recent subgroups.
Update from Ray Banks (via the Haplogroup G Facebook page)
The Big Y results seem to be getting out of hand. There are now 21 unprocessed files with 10 more that could be released at any time. I can only do several of these a week working without a day off. That means by the first of the month, there may be a 4 months supply. But there have been at least 10 more ordered during the current sale, which means more will arrive before the current inventory can be worked off.
Each time a new file is processed, there is a potential for a need for new targeted testing to prove new subgroups. So this adds to the problem, and to the cost of the targeted testing.
I mention this not to elicit sympathy, and it is difficult to assist. So the purpose is to explain a certain dysfunction that will necessarily develop. But many haplogroups have no such service, and others just work on small branches. This dysfunction can be ameloriated when multiple persons have sequenced within a subgroup because one can just compare results to pick out results for a particular location that exist.
05 December 2014 - Richard III
As you might have read in the newspapers, DNA testing on Richard III remains show that he actually belongs to haplogroup G2a, of which U1 is a branch. Although only a few markers were reported, it is possible to have an educated guess as to which subgroup he belongs to. U1 (in particular the DYS459a=7 subgroup Horton/Lesperance/Phibbs/Tims) and M406 are among the more likely candidates.
Arguments in favour:
- Mr Horton is the closest genetic match to Richard III in the FTDNA database on the markers provided
- The subgroup has on average 6/7 marker differences out of 19, but the common ancestor of these men and Richard III would have lived at least in 1400BC and likely centuries earlier, so this genetic difference seems reasonable
- Most markers are compatible with the U1 modal values (with a few exceptions - see below)
- His value for DYS533 is 10 which is rare below L140 (all existing U1 members who tested for this as part of STR111 have the value 9)
- His value for DYS389i is 13 which does occur in U1 - but not in the DYS459a=7 subgroup so far
If someone of the Horton subgroup is interested, it could be worth testing for DYS533. We only have 18 samples that have a value for this marker so it is not impossible that your subgroup is an exception within U1. Also, adding an ancestral location could provide clues. The one man in the subgroup who listed a specific English town, picked one that is only 35 miles from Richard III's birth place (although I admit this doesn't prove or disprove anything...)
We are hoping the team behind the DNA analysis releases more of the data so we can narrow down the possibilities.
22 October 2014 - Time to Common Ancestor
I have updated the status diagram (link on the results page) to show the estimated Time to Common Ancestor (TCA) for the more recent subgroups. This is based only on those men in each subgroup with 67 markers, so the accuracy and relevance may vary especially for the smaller groups. The tool used for the calculations is Y-Utility (http://www.mymcgee.com/tools/yutility111.html), and I have given the maximum value rather than the range. This means that some men within a given subgroup will have much shorter TCA's with their closest genetic matches.
17 October 2014 - BigY result for L1266
Update from Ray Banks (via the Haplogroup G Facebook)
A new Big Y from the U1 branch indicates that the two principal branches, L13 (European) and L1266 (western Caucasus) split about 10,000-11,000 yrs ago. The two groups only share one mutation besides U1, and the U1 man has 76 unique mutations.
This is actually my own BigY, and the SNP shared with L13 is Z6779. Ray further indicated that 76 mutations is a large number (I assume because there hasn't been any prior sequencing in that part of U1). It will be interesting to compare with Mikhail's BigY when available as this will hopefully result in further branching under L1264.
06 October 2014 - Branching under Z2003
The L13 part of the U1 project has seen a couple of interesting developments lately, so I thought it might be time for a state of play... (please also refer to the diagram on the results page)
We now have a number of men who tested L1263+, so I have been able to expand the possible scope of this SNP and mailed other men who are likely to share it. So far (and this is VERY speculative) one distinguishing feature of those confirmed and predicted through genetic distance to be L1263+ seems to be the STR value DYS447=22. This seems to work because all Z2017 men found so far have DYS447=23, as do all men confirmed negative for both. Mr Mowers has the value 24 but that is probably an additional mutation as it is very rare otherwise. IF the STR prediction proves correct then (non-Bentley) L1263 would be a very large subgroup, with an apparent geographical focal point in the northern part of continental Europe (northern Germany, Poland, Low Countries, etc...).
With regards to Z2017, I have now mailed the men genetically closest to the Hall family group to try and expand our knowledge about the coverage of that subgroup. This one is not easy to pick as we currently only have the Hall subgroup as a reference, and their distinguishing marker (DYS492=12) is not shared by any other L13 men so far. But as the experience with L1263 and the Bentley subgroup has shown, the markers of STR-oddity based subgroups are not necessarily a good guide to the remainder of men who share a particular SNP.
In terms of the existing STR subgroups, these have now all been confirmed as Z2003+ (refer entry for 3 October below) so it would be interesting to see whether they branch under L1263, Z2017 or neither (volunteers very welcome!). With the exception of a minority of men in the DYS534<15 subgroup, none seem to have the DYS447=22 typical for L1263, so initially it might make sense to test for Z2017 for all of them.
Finally, there are only a few men confirmed to be negative for all of the above so it is difficult to say much about this category. The only commonality I can see so far is that they seem to be concentrated around the southern half of the continent (Spain, Switzerland, Greece, Italy). Mr Del (who is likely to be in this category) has a BigY on order so hopefully we will be able to identify additional SNP's and learn more about this group.
Again, anyone who did not yet get an email from me but is interested in participating in our research, please email me and I will gladly provide a personalised recommendation...
03 October 2014 - Change to subgroup DYS446>18
Mr Wells' test has returned positive, confirming DYS446>18 as s a subgroup of Z2003. This was the last subgroup whose Z2003 status we were unclear about, and this confirms that all STR subgroups with Z2022+ have Z2003+ as well. In other words: the search for the elusive Z2022+Z2003- man continues...
I have re-ordered the project roster and adjusted the diagram to reflect this.
26 September 2014 - Tentative subgroup under L1264
We have identified a tentative new subgroup under L1264, possibly identifying a more recent branch of that SNP subgroup centred around the Circassian ethnic group from the Caucasus region in Russia. This subgroup is characterised by DYS389i=13 and DYS565=11. It currently consists of 4 confirmed men who tested positive for both markers, and 3 men who are predicted to belong to the group but only have 12 markers, hence their DYS565 status is unknown.
24 September 2014 - New Co-Administrator
Today we welcome Mikhail Yrievich Nochevnyy as Co-Administrator to the project. Mikhail is of Ukrainian/Russian background and has been researching the Caucasian L1266 group in particular. He is also the administrator of the Cossack/Козакъ project (https://www.familytreedna.com/public/Cossacks/default.aspx?section=yresults) which has significant overlap with the U1 Project.
19 September 2014 - L1263 Boundary testing
Up to now, we only had one man outside the Bentley family group (Mr Lowry) who had tested positive for the L1263 subgroup of Z2003. We always expected this subgroup to have a broader membership but had so far been unable to find other men with this SNP. We had therefore asked the men closest to Mr Lowry's markers to consider testing for it, and are pleased to report that Mr Hopper returned a positive result. We now have a much better idea of the STR patterns to look for, and this has enabled us to espand the L1263 (predicted) category with men who are genetically close to both Mr Lowry and Mr Hopper. Hopefully some will test for this SNP and further improve our knowledge about this subgroup.
18 September 2014 - NEW SNP Z2017 ADDED - Update from Ray Banks (via the Haplogroup G Facebook page).
It is almost a formality now that Z2017 will be added as a new ISOGG subgroup under Z2003 (within the U1 Project) Most of the U1 men in Europe are Z2003+. We have individual tests for Mr. Hall that show he is Z2017, and Mr. Bentley with Z2003 was negative for Z2017. Mr. Blackmon very near to Hall in marker values does not have his results posted, but he surely will be positive because Z2017 also appeared in a Mexican man in the 1000 Genomes Project who is not likely so near to Hall in marker values. So there is a green light now for Z2017 to be ordered by any Z2003+ man. And there were quite a number who tested for Z2003. This could be a very big group which hopefully divides the Z2003 men into 2 groups. The green light mentioned should be considered a red one if one has marker values near to Mr. Bentley. Consult with U1 administrator Alain Verschaeren before ordering.
17 September 2014 - Change to subgroup DYS459a=7
Mr Esperance's test has returned positive, confirming DYS459a is a subgroup of Z2003. I have re-ordered the project roster and adjusted the diagram to reflect this. My thanks to Mr Horton and Mrs Kelley for their help.
This leaves DYS446>18 as the only subgroup whose Z2003 status is unclear.
- Consists of the first two members of known Arab ethnicity with confirmed U1+ status (Mr Almandeel tested Z2003+)
- With only two members we cannot reliably establish the age or origins of the subgroup. Based on genetic distance the existing members have a 95% chance of sharing a common ancestor in the last 25 generations.
- The defining characteristic for the subgroup is DYS617=13. This appears reliable as an indicator as this value is not shared by any other confirmed L13+ men (at 67 markers) other than Mr Melik-Alaverdian.
U1+, L13+, Z2022+, Z2003+, (Z2017+), DYS492=13
- A predominately English subgroup consisting of the Hall and Blackmon family groups. Mr Jorgenson confirmed the subgroup’s Z2003+ status.
- Based on the current members, the subgroup is estimated to be about 1000 years old.
- The defining characteristic for the subgroup is DYS492=12. Again this appears very reliable as this value is not shared by any other confirmed U1+ men (at 67 markers)
- It is possible that this subgroup will be superseded by the Z2017 subgroup which is currently being validated (Mr Hall whose BigY triggered the discovery is a member), or it could become a subgroup of Z2017 if other men with this SNP do not share the STR oddity.
10 September 2014 - Project roster changes
I have now finished re-ordering the project roster to make it more intuitive to see how subgroups relate to each other. This is to be used in conjunction with the status diagram, which is available as a link from the "Results" page. If anyone has problems accessing this link, let me know and I email you the file.
04 September 2014 - Project update
It has been a busy couple of weeks with a number of developments which might be of interest:
- The orders to validate the tentative new Z2017 subgroup have now been placed, so hopefully we will have an update on its status in a month or so.
- The U1 project managed to quadruple its BigY tests during the end of summer sale. Besides Mr Hall's (which is completed), we have one more L13 man and two L1266 men who ordered the test, which will hopefully pay dividend in our research for new subgroups.
- Those of you who have looked at the project register in the last couple of days will have noticed two new STR subgroups (based on genetic distance and marker oddities rather than SNP's). If your name is in one of them, don't get too excited yet because they are tentative only and still subject to Ray's review. I will provide an update on their characteristics if and when they loose their 'tentative' status.
- I have removed the 'g2a....' references from the project roster, as they seem to become less and less relevant now that even FTDNA is using the SNP designations. If anyone misses them or wants to know what their 'old' designation is, you can still find them on the G-tree on the project website home page (https://sites.google.com/site/haplogroupgproject/). I am working on an improved diagram which will hopefully make it clearer where every subgroup fits within the project (tbc).
29 August 2014 - Update from Ray Banks (via the Haplogroup G Facebook page). We have contacted the men involved and are awaiting their reply to progress validation.
I have almost finished the BAM (raw data) file for Mr. Hall related to his Big Y testing. Family Tree in its analysis overlooked the fact he is positive for Z2017. This was found also in a Mexican man in the 1000 Genomes Project. We tested about 20 of the latter's mutations last summer in multiple men in Walk through the Y (WTY). This is the first time anyone else has been positive for that man's unique mutations.
So I have added Z2017 as a tentative new subgroup under Z2003 on the experimental tree. Mr. Bentley's L1263 subgroup would become the sister subgroup of Z2017. His WTY provides already all the info we need from that subgroup for the process of validating this candidate new subgroup. Hopefully we can gather what we need from Mr. Hall's testing and the man nearest to him in marker values (except Halls) whose tests will validate the new subgroup. Z2003 is the mother group, The Mexican man has Z2003, but Mr. Hall had but one positive Big Y read for Z2003 which is inadequate.
But this new subgroup seem much easier and cheaper to validate than most for which we have to provide validation.
The only other man within L13/Z2014 who ordered sequencing is Mr. Mortimor who only has 12 markers. 5 markers different in a second person are needed for a new subgroup, and this seems impossible to add his subgroup until we can find someone within Z2104 who mismatches 5 values of more. And there is no one within Z2014 who fulfills these criteria.
That leaves Mr. Verschaeren as the only man in U1 with a Big Y pending, and he is in an entirely different branch. Somehow I earlier overlooked he has a test on order
29 April 2014 - Update from Ray Banks (via the Haplogroup G Facebook page).
I am posting an e-mail just written to the G-U1 Project Administrator because near the bottom of the e-mail it uses a different methodology to approach the likelihood of finding useful results in testing (here within U1) in the new G tree.
It is extremely time-consuming evaluating each of the SNPs they added to the tree. After evaluating the first half of them, I posted on the Haplogroup G Facebook page that I urged caution in any ordering, and recommended instead concentrating on testing the near matches to the 4 G men who had one of the new items listed. None was within U1.
The additions they made [for you[ are within the Z2014 subgroup. The first item we can dispose of. It is PF6858. If you look on the master G spreadsheet, you will note it appeared in both the Z2003+ and Z2003- men in sequencing and also in Christy in Big Y. He is a subgroup of Z2014, but this was predictable since he is a Z2003+ man. It would be a shame to toss out all the work we have done getting Z2014 and Z2022 and Z2003 tested by so many men by switching to PF6858.
Page52 appears multiple times in this new G tree and also multiple times in other haplogroups. This is clearly an unstable item that should never have been added to the tree.
PF7506 was originally found in a haplogroup R man. Any of these new ones that appear in G will have originally appeared in another haplogroup because we have already identified the shared G items in the collection of the men who provided the PF, F and CTS series of SNPs. It is not rare that in a full sequencing each man will likely have a few SNPs found in other haplogroups also. If they appear in too many haplogroups, they will be useless because they are likely unreliable sites.
One can come up with a statistical likelihood of whether the G samples they used to draw the tree should have had a useful SNP. Because they are reporting results for 12,000 of the 20 million sites normally tested in whole sequencing, this reduces the chance of finding something useful popping up. Then one can take into account the percentage of G samples likely represented by G-Z2014 men. They looked at 50,000 Geno 2.0 tests. Abt 1500 were likely G, and 100 of those likely Z2014 and I think 6 of them are in the project. So what are the chances that they found a useful SNP among those 100 samples given the 12,000 sites tested? I also restricted this to the last 4000 yrs. I do not have a precise figure as with the other two major L140 subgroups, but that seems to be when there was a big expansion and breaking into lots of new lines. Normally abt 40 SNPs would be collected by ancestors during that time period. So we have to multiply 40 times the chance of finding a single SNP, and that comes out to a 2.4% chance. I will reduce this to 2% because some of these will be private SNPs. Multiply this 2% chance times 100 men and that is a 200% chance – or 2 new SNPs expected. That is better than many other groups because of the small amount of sequencing done.
So there is a good chance that PF7506 may be a valid SNP that occurs within Z2014. But the problem is that it likely occurs in a fraction of the 100 men tested. And none of them were among the 6 men who are in the project. So you may have to test 50 men in a shotgun fashion to find that one man with PF7506. And without knowing the marker values for PF7506+ men that really complicates things.
We do have 31 singleton mutations from the Z2014 man Mr. Hall. It may a bit more productive to have ySeq make up tests for some of these 31 and test the men near to Mr. Hall. ISOGG criteria will call for at least 5 markers different to escape the “private SNP” designation. There is one man with 8 markers different (Blackmon), then they skip to 15 or more markers different. In comparing just 37 markers with Hall, there are probably additional candidate men who do not have 67 markers. I am pursuing this approach within L640, and hopefully we can do the same for the Jewish Z724 group when a Big Y pending narrows the possibilities soon. It could be done in other subgroups where persons have had Big Y or Full Genomes testing