Delayed newsletter. This is the longest wait yet for a new edition of this newsletter.  But it was only in the last several days that quite a number of lab tests were reported that now make this edition of the newsletter worthwhille.  I have the impression that our wait may be tied to vacation time for lab employees.  There are some unreported lab tests, and if these are forthcoming soon, a second edition will be sent.  This will include a final installment on the Raetians, and I have promised an overview of Armenian G.

Just about all the news today involves SNP categories.

New Internet page. Though the data are not aligned yet, I have made available a working sheet that shows for the first time publicly the estimated numbers of 3 major G groups at a number of sites. Many of these were not from random population samples, and some of the data from the YHRD database could include samples from recent arrivals in the area  These data on this new sheet are overwhelmingly from Europe because these are only estimates, and there is very little testing available to confirm the SNP subgroup status of the marker samples outside Europe.

http://www.members.cox.net/generalbanks/geodistribution.html   I think the identification of the L497 (G2a3b1a2), L13 (G2a3b1a1a) men is fairly close to the mark, and using the DYS390=21 data as a marker for M406 (G2a3a) men this probably identifies about 90% of the M406 men.  The Swiss data here will be part of the final installment on the Raetians.

G Persons in the 1000 Genomes Project.  For several years the 1000 Genomes Project has been collecting selected DNA samples from around the world.  Learning about the findings from the Y-chromosome has been a do-it-yourself project.  Because this involves an area in which I had minimal familiarity, I have not been involved.  Ted Kandell has identified new SNPs from raw data in the past, but he seems to have been overwhelmed by other projects lately or had difficulty aligning the data.  Perhaps we can get him back on this is the near future.  However, with the help of GregM (his screen name -- I did not ask permission to use his full name) we have been provided some raw data files of use.   My involvement was prompted by the fact that persons from several other haplogroup projects not only reviewed the data but obtained lab numbers for the new mutations, and in recent days one (with the a Z prefix) was already added to the official Y-DNA halogroup charts at ISOGG.

So we are behind the curve on this. Some progress was made in the last week.  I learned how to comb through one type of data sheet provided by Greg and identified all the G samples, and I also added the haplogroup designation for about 150 other non-G persons to his spreadsheet.

https://spreadsheets.google.com/spreadsheet/ccc?authkey=CIOag_UD&key=0Agq_ez43qXCjdFlxemtlUnZ1Qk01cVhMRVBFcm5WX3c&authkey=CIOag_UD#gid=26

A lot of the data are listed on the bottom edge of the spreadsheet.  The purpose of identifying all the samples as to haplogroup was to see if there were more G persons.  I have a listing of the samples and haplogroups at

http://www.members.cox.net/generalbanks/1000genomesitems.html

The G samples are highlighted in green.  Afterworking with this data more, I perceive that we have identified all the G persons presently.  There will be more data later in the year.  The samples from western India may yield something.

In some ways, the finding of only 9 G persons out of perhaps 500 available is disappointing.  There are samples from some areas that have 5-10% G in the population and  where in the project the number of G samples is much lower than these percentages (Mexican Americans in Los Angeles, Medellin, Colombia, Puerto Rico, Iberia and Tuscany, Italy).  We were rescued by the samples taken in Utah where 3 are G men.  Perhaps sampling there included non-Mormon persons in the Salt Lake City area where there is a sizeable Greek, Hispanic and Jewish (actress Rosanne Barr) population.  The most surprising find was a G man in Beijing China samples.   

In the first phase of evaluating the data, I determined that all the men were G2, and five of the nine were G2a3.  L13 and L14 status was ruled for the men who had data on this. But these are only partial findings and not the full subgrouping of these men.

In the second set of data GregM provided there are about 12,000 SNP sites from the Y chromosome, with data specific to the 9 G men and 5 samples from other haplogroups for comparison.  The next hurdle is determining other known G mutations within the data.  The chromosome position listed in not the same as in the listed data for G SNPs.  I have not figured out yet how to translate this.  If anyone is experienced in reading data produced by SamTools or can assist in installing SamTools or browsers known to be able to read the published raw data, the help would be appreciated.

This process has potential for identifying quite a number of useful new SNPs.  Family Tree is leaving it up to haplogroup specialists to identify SNPs that might break down known categories because they can't create tests for all the SNPs.

The Y-Tree.  The full human Y-tree with distribution data I have been working on for a year has been updated with about 25 additional studies since last mentioned.  These include studies on haplogroup G, but there is nothing dramatic to report.  About 5 haplogroups have experienced major restructuring in recent months, and haplogroup A is in complete disarray because a researcher found that the next haplogroup down the line branches from within haplogroup A.

http://www.members.cox.net/generalbanks/completecategories.html

General Fund. Many thanks to those who donated to the General Fund since the last newsletter.  We have not spent much of this while waiting for the many pending lab tests to provide guidance for the next round of testing.

News by category and recommended testing.  Incorporated here are the results from two Walk through the Y tests reported last night, and the remainder of the L497 evaluation testing and other items.   Reference should be made to the project's five-page roster at http://www.familytreedna.com/public/G-YDNA/default.aspx?vgroup=G-YDNA&section=yresults  for your category, and info on ordering tests will be made at the bottom of this section.  Only one person within a family surname group needs to order SNP testing as all will have the same results. NEWS ITEMS ARE HIGHLIGHTED IN BLUE BOLD and larger fonts.

All G persons.  In the Walk through the Y Project tests for Mr. Fritsch and Bird, the lab reported a new L605 SNP.  This SNP is also available now for testing.  Mr. Fritsch is in a German subgroup of general L140 (neg for L497, neg for U1) and Mr Bird is in a subgroup of L497.   I do not know if this is a new area of the chromosome or not, but this SNP was not reported earlier.  I will be shortly placing an order for myself for L605 to see if the DYS568=9 group is also involved.  We need someone from the U1 or L13 group to also test and someone P303+ but negative for L140.

The same situation applies also to L402. This one has been around for a year from an uncertain source and is available for testing, but it was only in the last week that the lab posted it.  In a review of old Walk through the Y files, Mr. Haesenne, another L497 man also has this mutation in his Mar 2011 file and not in earlier files.  There was no documentation then that this was a G-only SNP.  I will be ordering L402 today myself to see if it occurs within the DYS568=9 group.  We need someone from the U1 or L13 group to also test and someone P303+ but negative for L140.

The Page00099 mutation which recently became available for testing is more of a challenge. This has been broadly tested in various G groups with everyone negative so far.  So we do not know if this is a private SNP or one covering a subgroup.  Any orders for this will add to our knowledge of this.

Predicted G1 (M285) men -- It would be helpful if the dozen men here whose relative has not tested to at least order M285 to confirm their G1 status if the full deep clade test is too expensive.  But the deep clade test will include testing for G1a which cannot be predicted except where I have listed men as being predicted for a G1a subgroup.

Predicted G1a and proven G1a persons (with G1a2, G1a3, G1a5, G1a6) -- Those not tested for the DYS494 marker.  This provides a significant division within G1a.  This marker does not apply to the men listed as G1a1 whose DYS494 status in known.  But the general G1a as well as G1a2, G1a3, G1a5, G1a6 men have yet to be tested for DYS494. 

G2 (negative for subgroups)  I have placed in this category tentatively until detailed testing is done our first participant from the nomadic Bakhtiari ethnic group of western Iran. His marker values do not match those seen in a research study for that group which has G1 persons among it as well as some seeming L13 men. The G composition of this ethnic group seems similar to other groups in Iran.   He is listed as Mr. Bakhtiari.  I cannot rule out either a G2a or G2c subgroup for this man who only has 12 markers presently.

G2a1a and its subgroups  -- Thanks to the efforts of Alan Berezov there have been upgrades to 111 markers for nearly all the Ossetian samples.  Earlier we had advanced testing for a grouping of a number of markers beyond the standard 67 only for the Athey family and for G2c men. http://www.members.cox.net/rayhbanks/ExtraMarkers.html but this is the first time we have seen G results for somewhat closely related men for markers 68 to 111.  It should be noted that there is less variation in markers 68 to 111 as compared to the other marker sets, meaning that these markers mutate less often.  When they do mutate, the possibility of a significant subgroup increases.  Because it would be helpful to have additional samples in other subgroups for comparison, there is yet no news to report.

G2a persons and those predicted G2a. Only one man listed in the roster is G2a but negative for all G2a subgroups.  So practically everyone in this section needs additional SNP testing except when a relative has had the testing.  Some men have tested and found negative for G2a1, (P16) G2a2 (M286) and sometimes G2a3 (L30), but hardly anyone among those with this testing have tested for G2a4 (L91) and G2a5 (L293).  The SNPs tested will show up at the bottom of your haplotree section.   When a lot of SNPs need testing, the deep clade test is the least expensive approach.

G2a2 -- We have word from 23andMe that one of our G2a2 men had a mutation in one of his DNA strands for the L223 mutation.  This may be promising news. We have multiple L223 tests pending at Family Tree, and it is hoped that L223 will incorporate all the men with double values at DYS19.  Some of these are L91 men.  If L223 proves to be a super group for both G2a2 and G2a4, they will become subbranches of L223.

Predicted G2a3 (L30) men.  We are unable to categorize these men adequately and only a deep clade test will remedy this for the several dozen men listed.

G2a3a (M406).  Due to the latest testing just reported, we now feel convinced that the new L90 SNP is confined to the L14 (G2a3a1) subgroup, and there is no need for M406 men to test for this.  However, there is great need for men only predicted M406 to order M406 or a full deep clade test. We can also predict with some accuracy who belongs to the L14 subgroup when someone has upgraded to 67 markers.

G2a3a1 (L14)  While some new L90 information is available, there are not enough men tested for L90 in the L14 subgroup to be certain that L90 and L14 occurred about the same time and are thus equivalent markers for G2a3a1. So far all L14 men also have L90, but if one of them is negative for L90, this would be the basis of a new subgroup. 

Only predicted G2a3b (L141)  There is great need for men listed in this category to have deep clade testing as this is just a general category for men who cannot be predicted in a more appropriate subgroup.

Only predicted G2a3b1 (P303)  There is great need for men listed in this category to have deep clade testing as this is just a general category for men who cannot be predicted in a more appropriate subgroup.  And those confirmed P303 but not tested for the L140 subgroup need to be tested.

Only predicted G2a3b1a (L140) There is great need for men listed in this category to have deep clade testing as this is just a general category for men who cannot be predicted in a more appropriate subgroup.   This testing will also determine the status for U1 (G2a3b1a1) which cannot be predicted for the men I listed in this category.

Only predicted G2a3b1a1 (U1) There is great need for men listed in this category to test for U1 unless a relative is confirmed U1.

G2a3b1a1b (Jewish U1 group).  Earlier we made this grouping a new subgroup, predicting they would be within U1.  The tests now show this is the case -- they are U1.

Confirmed or predicted L497 (G2a3b1a2)   The final targeted tests for this L497 SNP that was identified in January were reported yesterday.  There is fairly good evidence that L497 and the DYS388=13 group are one and the same.  All those men with DYS388= 12 but genetically somewhat close to DYS388=13 all proved negative for L497.  Thus those men predicted L497 on the project roster have a very high chance of being L497.   While additional data are always welcome, testing for the listed subgroups would be a more useful expenditure if you are in the predicted L497 group.  It must be mentioned that targeted testing was done of the L497 subgroups I established based on shared marker oddities, and G2a3b1a2 subgroups a, b. c, e, g, h and i all have the L497 SNP -- as predicted -- though some no longer have 13 at DYS388.  While the prediction for L497 is extremely likely to be correct, there are a handful of DYS388=13 men with only 12 markers who should test.  These men have (a) 12 at DYS393, or (b) something other than 11 at DYS392 or (c) something other than 22 at DYS390.  We also acquired a new member from the Persian Gulf area who becomes the closest DYS388=12 man near to DYS388=13 men.  Also, the L486 test is available for testing  by L497 men.  It forms a subgroup of L497, but only two L497 men other than the Flicks have tested for this.

Confirmed or predicted G2a3b1a2a1 (L42) There is great need for men listed in the predicted category to test for L42 unless a relative has tested.  Tests are pending for the large Bailey clain.  And for those confirmed L42 there is a new SNP available, L297, found in the Paver/Pever family.  We know that multiple Pavers have this, but no results for other L42 men.

G2a3b1a3c [A Jewish group within DYS568=9 group].   Earlier we reported that one man had a null value among the new 68 to 111 markers.  This seemed promising for a new subgroup. As it turns out all null values they are reporting in markers 68 to 111 (which are processed uniquely in Houston) must be considered suspect until fully investigated.  In this case the value was higher than the range they established and until this is corrected it will appear as null rather than as a blank.

G2a3b1a3g (new DYS568=9 subgroup).   This ao far all Hispanic subgroup has two 12 values for DYS464 and 23 for DYS390. 

G2a4 (L91) - predicted or confirmed.   Those only predicted L91 need to have a confirmatory test.  The new L223 SNP under evaluation we know will include L91 so no need to test for this.  However, those confirmed L91 need to test for the L166 and L167 subgroups as we cannot predict membership in that subgroup found among the Krauses.

G2a6 (double DYS19 value)  Those only who have not tested for L91 need to do so as half of L91 men are positive for L91.  No need to test if a relative has tested.

G2d  This a small group of G2 men negative for all other SNPs.  Because the split from other G men likely took place a very long time ago, it was hoped Mr. Jimenez from the G2d group would have many new SNPs when his 23andMe file was examined.  This was a sponsored test at 23andMe.  However, the results showed absolutely no new SNPs.

Ordering SNP testing.   The Family Tree DNA haplotree section where deep clade tests can be ordered is presently malfunctioning.  One has to instead choose Order Tests.  The Deep Clade Test is a standard test.  For individual SNP tests or individual markers, one has to choose instead Advanced Orders.   In the little window then choose either Y-SNPs or Y-STR markers, as appropriate.  Then type in the name of the test, following instructions to complete the order.

Discounts for upgrading to 111 markers. Ted Kandell has arranged for those who previously ordered markers beyond the 67 markers to receive a discount if they place their orders for an upgrade to 111 markers.   You must not have already ordered such an upgrade and must have certain minimum number of markers previously ordered.   Those who had such prior orders are listed at

 http://www.members.cox.net/rayhbanks/ExtraMarkers.html

Please do not contact me about this.   Someone will inform those eligible to see if they are interested in an upgrade. 

The Iceman is G2a4.  For some years scientists have been evaluating the intact, but mummified, remains of �tzi (the Iceman), a man who lived a little over 5000 years ago who was found in the melting icepack of the mountains between Italy and Austria.  He had on him quite a variety of man-made objects made of varied woods and metal.  Last month the head of the organization that is preserving his remains broke the news as to his Y-DNA haplotype.  This was likely a mistake to mention this because there is apparently a research paper pending publication which would normally be the source of the announcement.  The director's interview was broadcast on a German-language channel and unnoticed elsewhere until one of the viewers mentioned it in the comments on Dienekes blog.  Subsequently a clip of the interview in which he indicated the Iceman is G2a4 became available

http://dienekes.blogspot.com/2011/09/otzi-tyrolean-iceman-belonged-to-y.html

picture at http://dienekes.blogspot.com/2011/09/guess-otzis-y-dna-haplogroup.html

This is the third set of haplogroup G identifications from the 5000 yrs ago period  The one from northern Germany was found to be G2a3.  The Iceman in G2a4,   And the G2a skeletons in southern France were not checked for G2a subgroups.

G2a4 represents about 2% of our confirmed G samples, and is thus a rare subgroup within Europe.  Prior to the identification of this G2a4 sample, it has been difficult to get others to entertain the possibility that most G persons in Europe descend from a later ancestor who migrated to Europe.  The 67-marker info pertaining to our three largest European subgroups suggests that a later entry into Europe was the likeliest solution, but this latter info is not part of a published study which most persons want to see.  Because G2a4 is today so rare in Europe, this fact automatically proves that this ancient European G2a4 population did not succeed in leaving many descendants there.

Half of our G2a4 persons have double values for DYS19.   Because G2a4 (L91) has never been tested is a research study, we must rely on occurrences of a double 19 value as a proxy for the L91 test.  In this regard, double 19 is unusually more common in the Sardinian samples, and the director in his interview indicated the scientists found G2a4 samples from Sardinia.  That information is also news because I have not yet seen a confirmed L91 samples from Sardinia. The few verified L91 samples we have in the project are quite scattered geographically.  We know from a recent test that G2a4 is found within the Bakthiari of western Iran, and a double 19 was found among the Iranian Arabs as well.  One of our Indian men also has a double 19 value.

It should be mentioned also that they have been constructing a full genome for the Iceman.  This should provide useful additional information about G2a4 but without complete Y genomes from other G2a4 men far less info can be deduced.  Currently all our full genome samples are within G2a3.  His non-Y genome will have additional informaition because it will tell in totality what area of the world today the collection of SNPs are found.  We already know he comes from a rare mitochondrial group, and his branch is not otherwise represented today in available samples.

I must mention that the Dienekes site also has comments -- if you get that involved -- from a fellow who has been banned from multiple sites according to his own admission.  He is a prolific poster who makes personal attacks on anyone differing from his out-of-the-mainstream views.  With regard to G, he argues that it was Europe that G2a spread over the world, that the Etruscan island of Lemnos off Turkey was settled by Italians and the Etruscans were natives and not from the Near East.  With regard to the latter there have been quite a number of studies in the last decade of varying merit that suggest a Near Eastern origin for the Etruscans, and some of his postings are just a jumble of nonsensical things thrown together.  These postings are mentioned because I am sure to hear about them.

The 1000 Genomes Project.   Since our last report, I have been able to determine using a chromosome browser that one of the samples is L42.  And I was able to identify two addtional L497 subgroups using a G sample sequenced by Complete Genomics.  These are designated Z1824 and Z1823 and join Z725 and Z726 as L497 candidate subgroups.

However, all the rest of the news involving the 1000 Genomes Project is relatively bad.  There is no confirmation that Thomas Krahn of Family Tree DNA has ordered primers that would allow testing for any of these new G SNPs.  It seems like all new primers were halted at Z722 just before we became involved.  There have been about 50 primers ordered earlier.  He also has not added any of the many new shared SNPs with numbers higher than Z727 to his database.  So if you belong to haplogroups R, I, J and T you lucked out, but for the rest we may be locked out.   This could change, but I can only report that everything seems frozen as of back in July, and it is almost impossible to get Thomas to reply to anything, he indicating he has very little time left after doing his lab duties.

We know that the basic structure for most of our G subgroups is included within the list of SNPs found in only one person (singletons) in the 1000 Genomes Project.  We have L497, L42, L140, L640 and L13 men represented.  The number of SNPs found average over 20 per man, and some would be recent.  But some SNPs would be in the crucial 500 yrs -3000 yrs ago period.  The number of SNPs are far too many to merit primer development for each, but it was my thought that the approximate 6000--7000 SNPs occurring in all the 1000 Genomes Projects samples would be fertile ground for Walk through the Y or inclusion in an Illumina chip such as used at 23andMe.  So I contacted Bennett Greenspan, president of Family Tree DNA, to see if all these singletons were extracted, could he make use of them.  He replied in early August: "While I am not a liberty to make our future plans available to people outside the company I will tell you that these new SNP's are valuable to us and that we are exploring the bets (sic) ways to make the data available, therefore having the SNP's that you identify, even if they are probably singletons, is of value and we will be able to use the data for the eventual benefit of all in the community."

Based on this, I cancelled just about everything else I was doing to devote fulltime to extracting these SNPs.  The reports I have been using to gather this info provide various scores for the quality of the info.  The higher the score, the less likely this is something other than a SNP -- though one study showed the likelihood of being an SNP is still very high.  So I wrote Bennett again over a week ago to see if the lab would supply some scoring criteria for selecting the singletons.  He replied with a series of questions such as when would I finish and indicated he would talk to Thomas Krahn. 

Receiving no reply in a week, I wrote again, and he replied:

"I talked to Thomas once on this and he's not very interested in the singletons. Only snp's that have been seen multiple times. You should corner him as to exactly what he wants....Have you talked to Thomas on this? He really runs that show and I try to keep hands off unless things get out of control." 

I had sent Thomas a series of about 8 e-mails since July reporting new subgroup findings for all the various haplogroups I assembled -- about 100 new subgroups in total.  I received one very brief reply as the only reply to all these -- one limiting the number of subgroups that could be reported and unresponsive to other questions.  I immediately followed up to Bennett's directive yesterday to contact Thomas, but no reply has been received in 36 hrs.  Always in the past he responded quickly when responding.   So we have suspended all work on the 1000 Genomes Project.  It looks like all this may have been a total waste of time, but I have had  lots of other months even less productive in the past.  But perhaps the identification of these G singletons will still find a purpose.

Major new Caucasus study with G Information.   Yunusbayev and colleagues released a new study of the Caucasus region in recent days, titled "The Caucasus as an asymmetric semipermeable barrier to ancient human migrations," in Molecular Biology and Evolution.  Because of the concentrations of various G types in this region, a Caucasus study is always of interest.  The first major change from previous studies is that this includes Y-DNA, mitochondrial DNA and autosomal DNA all in one study.  The automsomal DNA is inherited randomly from both parents.  The major conclusion from this was that "Autosomal genome variation in the Caucasus reveals significant genetic uniformity among its ethnically and linguistically diverse populations, and is consistent with predominantly Near/Middle Eastern origin of the Caucasians, with minor external impacts."  I think it is theoretically possible to argue the opposite: that persons in the Caucasus populated the Near East/Middle East, but the colder periods of the Ice Ages made the Caucasus Mtn. area a bit too cold to sustain life and that the repopulation would have to have taken place from some lower altitude. With regard to Y-DNA, they re-used some earlier samples but added some new areas that were missing in the Russian study earlier in the year.  For the first time, we have samples specifically segregating South Ossetia from Georgia as it was unclear whether the earlier Georgian samples were taken in S. Ossetia then controlled by Georgia.  Still missing from any detailed study, however, are samples from Azerbaijan in the s.e. section.

Only G1 and G2a were tested.  For reasons unclear, not all the samples used previously were used for the Y-DNA report.  For example, samples from an earlier study used found about 15% G among the Mountain Jews (Tats) of the region, but in a reduced number in the chart they have no G listed for this group.  Although STR markers were tested, none were reported in the study.  The most striking finding was a split of 6 G men as G1 and 6 as G2a in Armenia -- 50-50 in other words.  The previous study by Nasidze a decade ago did not test for G1, but all G1 men have DYS392=12, and there was only 1 man among the 100 in Nasidze with 12 for DYS392, and he does not seem to be a G1 man.   In addition, in our own project, G1 men are in a considerable minority among Armenians.  It should be mentioned that the Cinniolglu study of Turkey almost a decade ago found the G1 in Turkey concentrated in the n.e. adjoining today's Armenia.  So there may be a pocket of G1 persons crossing the border into today's Turkey, but we do not know where in Armenia their 57 samples were collected.  And -- as always -- we would have liked to have seen the G broken down into more subgroups than G1 and G2a. 

http://dienekes.blogspot.com/2011/09/caucasus-revisited-yunusbayev-et-al.html  At the Dienekes site several of these charts are reproduced and can be clicked on.  The comments listed also indicate that they feel the shared ancestry of the various Caucasus groups is quite old, and they point out that the Turkish samples used for comparisons are from Cappadocia only -- which is in central Turkey. 

Update on New SNPs.  L223 which may cover both L91 (G2a4) and G2a2, has had samples at the lab since April with no results as they continue to work on the primer for the testing. 

I mentioned in the supplement recently that no one else had shown up positive for L640 other than myself.  This was also identified in a Utah man in the 1000 Genomes Project.  Consequently it was expected this would have some wide coverage within the DYS568=9 group.  Apparently this Utah man is somewhat closely related to me because everyone else is coming up negative.  Mr Pritchett  who shares an ancestor with me sometime in the Renaissance or Middle Ages time period agreed to pay for testing for L640 and his results might be reported as early as this week.  His results will start the process of seeing how back in time L640 goes.

We earlier reported L645 being found within a M406 (G2a3a) man.  Only one M406 person stepped up to order this, and we are waiting results.  L645 has been found in two of two M406 men tested. If there are no additional volunteers, we will have to tap into the General Fund to determine how widespread L645 may be.

There were also no volunteers for the new SNPs found within the P303+/L140- group and the L140+ (but not L497, not U1) group.  I attempted to contact the only man nearest to the P303+/L140- man to get him tested fro L668 and got no response.  There is a big gap to the next nearest relatives so I will keep trying with him. 

With regard to L660, L661 and L662 found within Mr. Labuz (L140+, L497-, U1-)  we have one sample in the lab who is genetically closest to Mr. Labuz being tested for these.  Likewise with L694 & L695 found with Mr. Zaleski (P303+, L140-) we have a sample in the lab being tested from the man nearest genetically to him. 

We have permission to test the man closest genetically to Mr. Hachigan, (L140+, L497-, U1-) but his new L697 and L698 SNPs are not available yet for testing.

The man from Lemnos Island off the n.w. Turkish coast was found to be L497.  In his 12 markers -- though DYS388=13 like the bulk of L497 men -- he has other atypical markers.  We are still waiting on his full 67 markers.  But this Lemnos find prompted some criticism for not paying for testing other men genetically near to L497 who are not DYS388=13.  The General Fund paid for testing two such men in the initial evaluation of L497.  To try to placate the criticism I have paid for testing of a DYS388=12 man genetically closest to the L497 group.  There are several dozen DYS388=12 men with 67 markers genetically closest to L497 but farther removed.  Let's see what the results are from this one man and from the Lemnos markers tests before wading deeper into. this

It should be mentioned that it has been tough getting volunteers coming forward for these new tests both in the G project and other projects.  Because most of our Walk through the Y participants recently come from targeted, poorly developed G groupings, nearly everyone had one or more new SNPs reported.  These could be common in some areas of the world and some will be restricted to individual families.  The volunteer problem is worse in some projects which have subgroups with dozens of new SNPs, more than most persons can afford at $29 each.  In addition in the last year, with the exception of the L640 orders, this newsletter and other information mailings only resulted in a few orders for SNPs and markers that define new subgroups.  I tried listing in one newsletter all the subgroups where additional testing was needed.  This resulted in two $7 tests within G1 so taking up stace to list additional testing needed probably was not useful.   We have had to rely on the General Fund for the most part to make progress in defining subgroups.

Also of note is the fact that project members continue to order less useful SNP tests than alternatives available.  Though every potential L497 person received a letter in the spring, for example, explaining they were highly likely L497 and a confirmed L497+ status would guarantee that everything in the deep clade test would be a certain finding -- despite this -- our predicted L497 men generally continue to order deep clade tests and ignore the less expensive and more informative L497. 

Calculating ancestry ages with whole genomes.   It is thought that the occurrence of a new SNP in Y-DNA is close to one per generation.  If this is correct, one can see how it would be possible to show how one is related to another person with some degree of relibility by having a listing of all the SNPs.  The problem is getting Y-DNA with enough passes of the data to guarantee that just about every SNP is identified.  This is quite expensive and usually involves several dozen passes.  We were able, for example, to identify about 20 SNPs equivalent to L497 in the 1000 Genomes Project and there are probably more in better mapped samples.  This shows that L497 existed in a small family group for perhaps 400 to 600 or more years depending on the duration of a generation in years.  The 1000 Genomes Project is giving intensive testing to some father-son pairs and even a few grandfather-father-son trios.  This will give us a better idea of the true mutation rate of the Y-DNA SNPs.  Having the complete genome info will also allow more accurate dating of the age of subgroups.  Presently we are relying primarily on the far less accurate STR marker mutations for which a number of conflicting time calculation formulas are available.

Final Comments on Raetians and L497 men.  Contining with information that may be pertinent to determining whether there may be a link between Raetians and our large L497 group of men.  It will be recalled that the Raetians were thought to be related to Etruscans who lived in north central Italy.   The Raetians who spoke a Raetian language, which some again think is Etruscan-related, lived principally in eastern Switzerland, southern Germany and western Austria.  They eagerly became Roman soldiers.  In Britain, the L497 often have Swiss L497 men among their closest matches though each is from a different branch of L497. 

My final comments relate to our L497 men in Switzerland.  I observed earlier that there were not many L497 samples from eastern or western Switzerland.  They are primarily central Swiss. The Alan Samaritans were known to have settlements in western Switzerland and the Raetians were most associated with eastern Switzerland.  I theorized that central Switzerland might have provided more emigrants to North America where most test participants live today and this could explain the distribution.  To test this theory, I checked some other large Euorpean haplogroups and found that eastern Switzerland, in contrast, was well represented in their samples.  I do not know if this is consequential or not.  There is a long open valley from western to eastern Swtizerland, and there is typically some movement back and forth for certain distances within stationary populations. The total width of central & eastern Switzerland is about 60 miles (100 kilometers), but some of the mountains in the east are not habitable. One can evaulate this new information as one wishes.

General Fund.  Thanks so much to those who made contributions of the General Fund since the last newsletter.

Country of the Month: Armenia

This feature has not been used in recent months, partly due to lack of additional countries with adequate samples and partly due to lack of time.  The feature is being resurrected this month to fufil a commitment to cover Armenia.  This country is south of Georgia and Azerbaijan which border the southern portion of the Caucasus Mountains.  Iran is today on the south with part of Azerbaihan, and Turkey is on the west.  The Armenian kingdoms once spread beyond the current borders and considerable numbers of persons with Armenian surnames are found in multiple parts of Turkey.  Minorities within Turkey have had an uneasy relationship with the Turkish governments.

New Subgroups.  There are multiple new subgroups to announce.  In total we are getting close to having 100 identifiable subgroups within haplogroup G.   Some new ones are based on new SNPs, and others are based on shared marker value oddities.  As always, new subgroups based on shared marker value oddities require almost always that members of the subgroup have each other as nearest matches when comparing marker values.   Instructions for ordering the items involved as listed at the end of the discussion.

1.  G2a3a2 - L645+   There are so far two Armenians and a Swiss man positive for this new SNP.  It is the first SNP subgroup added to G2a3a within about five years, and it seems about as significant as the L14 subgroup in helping to subdivide G2a3a persons.  Based on some cluster analysis done on our 67 markers by Georgios Kechagioglu and a noticed shared marker value, it seems we may be able to predict who belongs to this subgroup.  Indications are that men who have a 9 value at DYS578 belong to this new L645 subgoup.  If this is so, the subgroup is quite varied geographically.  We have a few more L645 tests pending to better confirm the relationship between DYS578 and L645.    It should be mentioned that our G2a3a3a5 - DYS392=12 subgroup also has 9 for DYS578, but they are not genetically close to the L645 men and thus likely negative for L645.

So anyone in the G2a3a (L406) group who has 9 for DYS578 and is also NOT listed as part of the G2a3a3a5 subgroup would be eligible for testing for L645 with some likelihood of being found L645+.  The L645 subgroup is now listed in the ISOOG G tree and is accepted as an official category.

2. G2a3b1b - L694+  This subgroup is also now accepted at ISOGG as an official G category.  We do not have a large number of P303+ men who do not belong to the monster L140 subgroup of P303, but here we have a subgroup within P303 which incorporates some of such L140- men.  Both Mr. Chesky and Zaleski are positive for L694, but men at a larger genetic distance are negative.  So this may be an all Polish subgroup.  There is no reason for anyone within the project to order this SNP at present since one has to be genetically close to these two Polish men.  It should be mentioned that Mr. Zaleski was also positive for the new L695 SNP, but Chesky is negative.  So we will treat L695 as a familial, private SNP.  There is additional news for P303 men negative for L140.  Our second Indian participant in Walk through the Y has been found positive for the new L769 and L770 SNPs.  Our first Indian participant who is in this same category was not tested for these, and he has ordered these to see if L769 and L770 have wider coverage than just one man.

3. G2a3b1a4 - L660+, L662+.  This subgroup is likewise now accepted at ISOGG as an official G category.  The Polish man, Mr. Labuz, was positive also for L661, but L661 seems restricted just to his family and will not be used.  At present we do not know how widespread this L660 group may be.  Those persons who are G2a3b1a (L140+)  should consider ordering L660/L662 UNLESS they belong to the following major L140 subgroups (or are predicted to belong to them) which we know are not involved with L660 and L662:

(1) U1/L13 -- G2a3b1a1 ....(2)....L497 -G2a3b1a2 ....(3) ... DYS568=9 - G2a3b1a3  None of these men should order L660/L662.

4. G2a1a1b - DYS 391=11.   G2a1a is incredibly common within the Ossetian population at the center of the Caucasus Mountains.  We had already established a subgroup of G2a1a for those Ossetians and others who have 19,21 for YCA.  This was further subdivided, adding a sub-subgroup of men who are DYS438=9 and DYS391=9 (the most common Ossetian grouping).  But there is yet a second sub-subgroup of YCA 19,21 men being announced here -- men who have DYS391=11.  You should quickly notice that this is 2 values different from the men in the other sub-subgroup, suggesting there was a two-value mutation at this marker which is not all that changeable.

5. G2a3b1a1c - DYS537=10 & DYS455=10.   This is a new subgroup within U1 persons. Most U1 men belong to the big European L13+ subgroup, but this is a much smaller European subgroup.  We note here some similarities in coverage to L497 clusters where English and Swiss men seem to share a common ancestor within the last 2000 yrs.  This U1 subgroup may have been imbedded among the L497 men.

6. G2a3b1a1d - DYS531=12.  This is yet another U1 subgroup but with a concentration in the Turkey-Armenia-Iran triangle.   Within U1, the value for DYS531 of 11 is found almost exclusively.  The members of this subgroup are the exceptions.

7 & 8. G2a3b1e and G2a3b1f.   For those persons P303+ but negative for L140, we indicated above there is a new SNP subgroup with limited coverage, but there are two additional subgroups based on shared marker value oddities.  DYS594=9 and DYS459=8,9 is so far an all Hispanic subgroup.  And DYS392=12 and DYS520=20 is quite varied in make-up.  We normally would not use DYS520, but this is several values lower than other available samples.

9. G2a8.  DYS481 less than 18.  This is a new G2a group which seems most common among Italians.  The values below 18 represent multi-value jumps.  Some uncertainty about relationship of this new subgroup to some other G2a subgroups must remain.  It shares the 10 value at DYS392 with the G2a7 group.  So it may actually be a subgroup of G2a7 or vice versa.  And we don't know how G2a8 men would stand with regard to L223 which may bring together a few G2a subgroups.

10. G2a9.  DYS389=9 and YCA=19,20.  This is a very Armenian subgroup.  The only relationship not known about these men is the status with regard to L223.  We have representative samples in the lab which have had the first L223 results pending since April.  But the lab still has not resolved the problems with the test specifications for L223.

To order new SNPs, go to your results page.  Choose Order Tests.  Select Order Advanced Tests [not Advanced SNP tests].  In the little window on the left choose SNP.  In the window on the right, type in the name of the SNP, such as L645.  Then hit the Find button.  When the info appears, choose Add, then follow the instructions to payment.  The cost is $29 each US. 

One can follow the same directions to order an individual marker, but instead STR marker is chosen, followed by entering the name of the marker.  The cost is $7.20 US for most items. 

A $10 transfer fee is charged if one of your swabs is not already in Houston.

1000 Genomes Project Update.  There is little new to report.  The lab did add to the description of two of the SNPs I discovered in the data, information about primers, which are needed to construct the tests.  However, these seem to be from an existing database of primer info, and in no way guarantees that primers will actually be ordered.  There is an upcoming conference for project administrators in Houston, and we should know by the next newsletter whether the lab will offer some new product that will take advantage of the G SNPs found in the 1000 Genomes Project.  One paper published about the 1000 Genomes Project indicates they found 40 million genetic variations in the samples (if I remember the figure).  But whether a number of these are just false positives is unclear.  For sure, they did not individually sequence each of these to confirm the number.

23andMe New Product.  23andMe has recently offered for sale to a limited number of customers a test which will produce results for a large number of locations within exons.  Exons are the coding portions of genes.  Perhaps about 10% of the SNPs about which we know also occur in such locations.  So this product might be the most useful one available to us if it gives information on a large number of SNP sites unless Family Tree comes up with a better product.  They are charging $1000 for the experimental version of this, but I would expect the price of a product offered to customers to be less.

The General Fund.  The General Fund has partially helped to identify some of the new categories, and we will need General Fund money to determine the boundaries of L223 and the 1000 Genomes SNPs if they become available.  Many thanks to those who donated to the fund since the last newsletter.

�tzi,the Tyrolean Iceman, our G2a4 cousin, is back in the news.  Dienekes is speculating that the issue of Science may be devoted to his genetics, and a mummy conference will take place in the next few days.  The Iceman who is a little over 5000 years old and was frozen in the Alps for that entire period, had brown eyes and hair and was lactose intolerant -- an odd finding for a shepherd.  This info is from the promo from the 26 Oct edition of the American science tv program Nova.  They are saying the Iceman has closest connections to persons living in Sardinia, Spain and Italy and not to persons in Africa or Asia.  But it is unclear if his Y-DNA was compared to G2a4 men from these latter areas.  Nova has made its old programs available for streaming, but I could not predict when this one will be available as I think there is a delay.

My prediction at the beginning of 2011 that 2011 would not see much progress in defining G categories has turned out to be quite incorrent.  I felt that 23and Me and Walk through the Y would not have dramatic breakthroughs because they had gathered most of what could be gathered.  There was no hint that the 1000 Genomes Project would have a wealth of new things and that the labs were moving toward much more thorough testing of the Y chromosome.  I guess prudence dictates it is best nor to predict anything for 2012.

There are new tests and new findings in a number of G groups, and it is best to cover them one by one.  It is probably best also that you know which category in which you are listed or predicted to be part of.  These are found in the project roster.  

http://www.familytreedna.com/public/G-YDNA/default.aspx?section=yresults -- reset page size to 3000 or page through page by page if you do not have much computer processing memory

The discussions below will presume you know which is your category.  There has been some restructuring of the categories, as will be explained.  Wherever, a new test is mentioned, the instructions for ordering each will be listed at the end of the news.

The DYS568=9 group (G2a3b1a3). During the fall, a number of project members who are in the large DYS568=9 group tested for the L640 SNP.  Because a random, anonymous Utah (USA) man in the 1000 Genomes Project had L640 and one person in the project had L640, it was thought L640 would likely encompass many DYS568=9 men.  It turned out L640 is instead found in a small English subgroup of DYS568=9.  Now this same anonymous man in the 1000 Genomes Project has provided us another new SNP, namely Z1903.  We were lucky to have this test.  This anonymous Utah man has about 20 new SNPs, but the lab will not provide tests for these because there is not another DYS568=9 man in the 1000 Genomes Project sharing the Utah man's SNPs.  If the lab provided tests for every SNP found in 1000 Genomes there would be about 7000 new SNPs, and they can only provide tests for the most promising of these.  In the case of Z1903, however, the mutation is only a few positions down from a mutation for which they already have a test.  So they were able to use the existing test primer to run also Z1903 without any new expense to the lab.

I am a L640+ man just like the Utah man.  When Z1903 became available a few weeks ago, I tested for Z1903, and I also have the Z1903 mutation according to the results.  At this point we do not know if Z1903 is found only within the L640 subgroup or whether it may be the long sought SNP to define the DYS568=9 group.   I have permission for the upcoming Wednesday batch to test two men who are most likely to give a preliminary estimate of how widespread Z1903 might be.  One is a L640+ man who is farthest from me genetically within the L640 subgroup.  The other is a man closest to the L640 group but who was found to be negative for L640.  But if any other person within the DYS568=9 group also wants to test for Z1903, this will provide additional information.

The G1 men negative for G1a.  We reported in an earlier newsleter that 8 SNPs were found in a man negative for G1a who was also the last participant in the Walk through the Y Project.  The 1000 Genomes Project info allowed us to determine that two of these are found in all G men.  We now have some additional info on five of the six new G1 SNPs.  One of these, L833, is now known to be present in all G1 men and os thus of not much interest.

However, we also now know that L830, L832, L834 and L835 are absent from G1a men.  L831's position within G1 is still uncertain.

It happens that the man who had all these SNPs (in Walk through the Y) is assigned in the project as a subgroup I created, designated as G1c.  G1c men share a mutation to 12 at marker DYS446.  And we now have two men within the G1c subgroup who are positive for all these new G1 SNPs.  So the G1c subgroup (see our project roster) are not the prime targets for further definition of the boundaries of these new SNPs within G1.  What is needed now is persons within the G1 group who are negative for G1a and G1b and who also lack the 12 value at DYS446.   You would be testing L830, L832, L834 and L835 -- and may be needed for L831.  We have been lucky so far to have volunteers for all the testing on this new G1 SNPs.   It will be expensive to use the General Fund to test for five SNPs simultaneously. 

The results so far meet all the requirements for a new G1c category defined by SNP(s).  However, ISOGG

(see http://www.isogg.org/tree/ISOGG_HapgrpG.html ) with the cooperation of the involved scientists is moving toward removing any SNP categories that do not meet some minimum criteria for group size.  The scientists have provided us with with two G categories that seem based on a single person's results.  G1b may be such a one-man category.  If for no other reason, we also need a particupant from each category so we can test relevant new SNPs on him.   G1b definitely was created by the scientists and is rumored to be based on a single person in Sri Lanka and we have no participants in the G project who are G1b.  The ISOGG person who is the contact person who corresponds with the scientists has written Dr. Karafet to try to determine whether G1b is just a one-person SNP.  If so, then G1b may be eliminated from the ISOGG tree, and replaced by the new SNPs with the L830's numbers.  But by the next newsletter this should be resolved.

The G2a3b1a2 group (L497 men).  There are several new tests available for L497 men (the great bulk of whom are also DYS388=13).  About 40% of the men in the project belong to this L497 group. 

1.  The first SNP was found in the 1000 Genomes project and is designated Z725.  In the 1000 Genomes Project, this was found in L497+ men from Utah (British Isles ancestry), Puerto Rico, and Tuscany, Italy, but absent in a Mexican American.  Based on this, the coverage for this new SNP would seem to encompass most L497+ men not not some.  But the Mediterranean dominance of the L497 men in 1000 Genomes may be misleading about this.  The Utah Z725 man here is also part of the L43/L42 subgroup, and this guarantees that all L43/l42 are positive for Z725.  There is thus no need for L43/L42 men to test for Z725.  Any L497 men are welcome to test for Z725 to determine its boundaries.  There were several other SNPs that seem more promising to L497 men, but the lab was not able to provide primer tests for these.  The General Fund is submitting the Chletos sample for Z725 testing.  He is a L497+ who seems most irregular in his marker values and may be an early branch from L497, which is what Z725 seems to be.

2.  We were able to get a new test for Z1901 which was found within a 1000 Genomes L497 person -- This test is accidentally available because the mutation is quite near to another mutation which has an existing test.  Because it is found in only one of the L497 men in the 1000 Genomes project, it is very likely it does not have wide coverage.  This one Z1901+ man is from Tuscany, Italy.  It is difficult to target who should test for Z1901.  I chose two men from Italy for the first two tests.  One is negative for Z1901, and the other is not yet reported.  The L497 men from Italy are poorly tested, and they do not always indicate where is Italy was the ancestral home.  But Z1901 may extend beyond Italy.  The Z1901 man from Tuscany is from a subgroup from which most L497 derived.

L140+ men (G2a3b1a) who are neg for the U1/L13 and L497 subgroups.  Another new SNP test available today is Z724.  This will probably fill in the last major hole in the haplogroup G tree encompassing those men who are L140+ but who were found negative for U1/L13 and L497.  Not many persons have tested for L497, but I think I have all L497 men listed in the L497 section of the roster who might be L497.  It is the plain G2a3b1a men on the roster who need to test for this, as well as those I have listed as belonging to G2a3b1a4, G2a3b1a5, G2a3b1a6,G2a3b1a7 and G2a3b1a8..  The latter are categories based on shared marker values, and we need to know where they stand with regard to this new Z724 SNP.

In the 1000 Genomes Project, this Z724 mutation is found in a L140+ man from Beijing, China, as well as a man from Utah, USA.  The latter belongs to the DYS568=9 (G2a3b1a3) group and logically Z724 will be found in all DYS568=9 men, but we do not know much about this Chinese man.

"Old" G2a2 and G2a4 and G2a5 categories abolished.  In the project's listing of categories, you will find yourself in new categories if you were found positive for M286 (now old G2a2) or for L91 (now old G2a4).  Since April, we have had results pending for all the men who ordered L223.  There were enough results reported earlier in the month, that it can be confimed that we are now listing L223 as the new G2a2 category.  As such, it becomes the mother group for M286 -- now G2a2a -- and also for L91 -- now G2a2b.  L91 has several subgroups, and these have been moved underneath G2a2b. 

I have also moved those men who have double values for DYS19 into the predicted L223 category. It seems clear that L223 encompasses all such men whether in the M286 and L91 subgroups or in neither. I also eliminated G2a5 and renamed it G2a4 in order to fill the vacancy caused by moving L91.

There is some interest in this L223 finding from the research community because you may recall that the Iceman mummy was revealed this summer to be G2a4 (now G2a2b).  And his nearest relatives were said to be in Sardinia.  Sardinia has an above average percentage of men with double DYS19 -- who we now know are likely L223+.

These changes have not yet been reflected in the ISOGG charts or Family Tree DNA's charts.  It will be necessary to show that all the subgroups parallet to L223 are negative for L223.  This was actually already determined in 23andMe testing, but they will want individualized sequencing to confirm this before changing the charts.  I have General Fund money reserved to test L223 for whereever needed to prove its position, but I was reluctant to use the money for this until they could confirm they could do L223 testing.  There are still a number of L223 results unreported.

Ordering New Tests.  To order any of the new SNPs at Family Tree DNA, on your results page choose Order Tests.  Then choose Order Advanced Tests (not Order Advanced SNP tests).  Then in the next screen, choose SNP in the little window.  Type in Z1903 (or whatever other test) and hit the Find button (hitting the Enter button will not work).   Then hit Add to add the item to your shopping cart, and proceed to payment.   The selection process has to be repeated for each item, and I urge great care in typing the number of the test as others have sometimes typed in wrong numbers.

Thanks -- the General Fund.  Many thanks to those who donated to the General Fund since the last newsletter.  Now that we have all these new SNPs we will be spending more than average to determine which subgroups are included in the new categories where volunteers are not forthcoming.

The Roche 454 Sequencer and whole genome testing.  We reported in the last newsletter that Family Tree DNA is testing the Roche 454 with the intention of making a test available which will provide results for large parts of the Y chromosome.  There is nothing new to report, but it is known that one or more of the attempted runs did not produce useful results.  They may possibly need blood samples -- rather than sputum samples.  However, some of the other genome tests work well with sputum.  We are actually very much in the dark on all this.  There is a Chinese company BGI (Beijing Genomics Institute) 

http://www.sfgate.com/cgi-bin/article.cgi?f=/c/a/2011/11/30/BUJR1M5INJ.DTL

that will do a complete scan of the Y DNA (and other DNA segments) for $500 using blood samples.  However, they require a large number of orders at this price.  So it still remains a question mark as to whether anything practical will be available for us.

Recent General DNA and Archeological News
The most recent population genetics studies summarized at Dienekes blog:
http://dienekes.blogspot.com/
include these items [with posted comments]:

   (1) On 16 November he discussed the Armenian Y-Chromosome article from Herrera et al.  with some discussion of G2 in general

http://dienekes.blogspot.com/2011/11/armenian-y-chromosomes-revisited.html