PDA

View Full Version : Is there a y-dna link between the Yamna culture and R1a?



epp
02-08-2017, 11:12 PM
I understand that nearly all known Yamna samples are R1b, so why do people mainly identify Yamna with R1a?
There seem to be no signs of early Ukrainian/South Russian R1a. The earliest traces of R1a are Middle Eastern, and the earliest branches from there were North Western Europe (R1a1a1a), the Baltic (R1a1a1b1) and the Middle East (R1a1a1b2). Perhaps the first R1a group estimated likely to be of Ukrainian/South Russian origin is YP314, which yfull TMRCAs as only 2,400 ybp (long after the Yamna period ended).
All the earliest R1a branches seem to have migrated West to East (R1a1a1a into Scandinavia, R1a1a1b1 into Russia, R1a1a1b2 into India) - the opposite direction to Yamna, which is said to have spread East to West into the Corded Ware Baltic regions.
Any thoughts?

lgmayka
02-08-2017, 11:35 PM
There seem to be no signs of early Ukrainian/South Russian R1a.
Genetiker decodes (https://genetiker.wordpress.com/2017/02/04/y-snp-calls-from-mesolithic-and-neolithic-latvia-and-ukraine/) Ukraine_N1 (4519–4343 B.C.) to be R1a-M459*.

rms2
02-08-2017, 11:42 PM
I understand that nearly all known Yamna samples are R1b, so why do people mainly identify Yamna with R1a? . . .

I have not heard of anybody doing that these days. They identify Corded Ware mainly with R1a.

However, I do expect to see some R1a in Yamnaya, as well as some R1b-L51. No y-dna results from western Yamnaya have been published yet.

Coldmountains
02-09-2017, 12:01 AM
I understand that nearly all known Yamna samples are R1b, so why do people mainly identify Yamna with R1a?
There seem to be no signs of early Ukrainian/South Russian R1a. The earliest traces of R1a are Middle Eastern, and the earliest branches from there were North Western Europe (R1a1a1a), the Baltic (R1a1a1b1) and the Middle East (R1a1a1b2). Perhaps the first R1a group estimated likely to be of Ukrainian/South Russian origin is YP314, which yfull TMRCAs as only 2,400 ybp (long after the Yamna period ended).
All the earliest R1a branches seem to have migrated West to East (R1a1a1a into Scandinavia, R1a1a1b1 into Russia, R1a1a1b2 into India) - the opposite direction to Yamna, which is said to have spread East to West into the Corded Ware Baltic regions.
Any thoughts?

Earliest R1a found yet was from Mesolithic Karelia which was EHG with basically zero traces of Middle Eastern ancestry. Also Dnieper–Donets was also R1a just like Khvalynsk which was both R1a and R1b. Yamnaya is irrelavant for R1a and was likely not PIE anymore. R1a is from pre-Yamnaya steppe cultures which were likely more diverse in different R1a/R1b lineages. Most of them died out when some clans became powerful and pushed out other steppe clans.

rozenfeld
02-09-2017, 12:09 AM
There was a leak, that one of the samples from Yamnaya, analyzed by Reich's lab, had R1a. The sample is however not yet published. So I will wait for next publication.

Michał
02-09-2017, 12:25 AM
The earliest traces of R1a are Middle Eastern
I am not aware of any such findings. Could you please provide a reference for this?
Also, the early diverged subclade of R1a named YP4141 (or R1a2) is found both in Europe (so far mostly in Western Europe) and in the Middle East. YP4141 is definitely most frequent in the Middle East, but this practically includes just one subclade named YP5018 with the relatively young TMRCA age of about 3000 years, so it could have been brought to this region quite recently, ie. in the historical times.

Michał
02-09-2017, 12:38 AM
Perhaps the first R1a group estimated likely to be of Ukrainian/South Russian origin is YP314, which yfull TMRCAs as only 2,400 ybp (long after the Yamna period ended).
I am wondering where this comes from. I definitely wouldn't say that the modern distribution of YP314 makes it much more likely to have originated in Ukraine. Actually, it seems that YP314 is less likely to be of Ukrainian origin than many of the closely related clades, like L1280 or Y2902 that, by contrast to YP314 (or to its parental clade S18681), are present not only among the Eastern and Western Slavs, but also among the Southern Slavs.

Generalissimo
02-09-2017, 01:02 AM
I understand that nearly all known Yamna samples are R1b, so why do people mainly identify Yamna with R1a?
There seem to be no signs of early Ukrainian/South Russian R1a. The earliest traces of R1a are Middle Eastern, and the earliest branches from there were North Western Europe (R1a1a1a), the Baltic (R1a1a1b1) and the Middle East (R1a1a1b2). Perhaps the first R1a group estimated likely to be of Ukrainian/South Russian origin is YP314, which yfull TMRCAs as only 2,400 ybp (long after the Yamna period ended).
All the earliest R1a branches seem to have migrated West to East (R1a1a1a into Scandinavia, R1a1a1b1 into Russia, R1a1a1b2 into India) - the opposite direction to Yamna, which is said to have spread East to West into the Corded Ware Baltic regions.
Any thoughts?

R1a is first attested in the ancient DNA record in far Northeastern Mesolithic Europe. Then also very likely in Neolithic Ukraine and certainly in Eneolithic Southern Russia. Not a single ancient Near Eastern sample belongs to R1a.

Expect plenty of ancient R1a from all periods in the steppes north of the Black Sea.

Don't expect any R1a in the Middle East until the Bronze Age at least.

parasar
02-09-2017, 02:16 AM
...
There seem to be no signs of early Ukrainian/South Russian R1a. ?

1000bc or older.
Don/Tanais R1a
http://www.ssc-ras.ru/files/files/Southerns%20Regions%20Population.pdf

epp
02-09-2017, 09:00 PM
Thanks for all your many pieces of information!

Parasar - the 1000 BC samples seem too recent to be Yamna, but are interesting nonetheless.

Generalissimo - I haven't seen convincing evidence yet, although your expectations re- the Southern Steppes were probably similar to mine.

Michal - your identification of L1280 & Y2902 as possible Ukrainian-origin subclades is interesting, not least because yfull estimates their TMRCAs to be 200-300 BC (very much post-Yamna). FTDNA's large R1a database has quite a few M417- samples that appear to be Middle Eastern (e.g. Iraqi, Yemeni and Jewish).

Coldmountains - do you have references for the information you have given? (I agree, there might have been various R lineages in the Steppes. This was my guess for a current R1a point of origin, although (as you suggest) many (or perhaps all?) of these lineages died out. I cannot see evidence yet that they are the most recent common source of surviving R1a lineages.)

Igmayka - Genetiker's "calls" are interesting if we can rely on them - I note the 4,500 BC Ukrainian apparently-M459* sample also had a positive call for a subclade of YP4887 (which yfull estimates only came into existence 363 ybp), so Genetiker's assessment doesn't seem assured. In fact, yfull estimates M459 itself to have a TMRCA of only 3,500 BC, and my view is that yfull generally overstates TMRCAs. Also, how easily is this squared with the FTDNA database, in which all East European samples appear to stem from M417+, and all M417- samples look Jewish and Arabian in origin?

Megalophias
02-09-2017, 09:26 PM
Genetiker's "calls" are interesting if we can rely on them - I note the 4,500 BC Ukrainian apparently-M459* sample also had a positive call for a subclade of YP4887 (which yfull estimates only came into existence 363 ybp), so Genetiker's assessment doesn't seem assured. In fact, yfull estimates M459 itself to have a TMRCA of only 3,500 BC, and my view is that yfull generally overstates TMRCAs.

Y-Full's TMRCA for M459 is 12 500 - 15 800 years. Are you talking about M417?

Genetiker's analysis includes a few false positives and negatives, which is natural (sequencing errors, ancient DNA damage, or recurrent private mutations). Other samples have the same.

Anyway, we now know that basal R1a (and R1b) was pretty common in Mesolithic Eastern Europe, yet seems to be very nearly or entirely extinct today. It seems to me that when it comes to rare basal lines today, as with ancient DNA, positive evidence counts for something but negative evidence doesn't mean much. So the basal R1a in West Asia needs explaining, but lack of basal R1a in modern Eastern Europe is just due to lots of replacement.

lgmayka
02-09-2017, 11:17 PM
my view is that yfull generally overstates TMRCAs.
The more common view of experts in this forum is that YFull (slightly) underestimates TMRCAs.


Also, how easily is this squared with the FTDNA database, in which all East European samples appear to stem from M417+, and all M417- samples look Jewish and Arabian in origin?
Your statements are false. Take a look at the first page of the R1a Project spreadsheet (https://www.familytreedna.com/public/R1a?iframe=yresults).
M420+ M459- : British Isles, Italy
M459+ M198- : Poland, Belarus, Czech Republic, Italy, Russia
M198+ M417- : Germany, Netherlands, Puerto Rico, British Isles, Austria, Macedonia

epp
02-09-2017, 11:27 PM
Y-Full's TMRCA for M459 is 12 500 - 15 800 years. Are you talking about M417?

Yes, sorry - M417, not M459.


Anyway, we now know that basal R1a (and R1b) was pretty common in Mesolithic Eastern Europe, yet seems to be very nearly or entirely extinct today. It seems to me that when it comes to rare basal lines today, as with ancient DNA, positive evidence counts for something but negative evidence doesn't mean much. So the basal R1a in West Asia needs explaining, but lack of basal R1a in modern Eastern Europe is just due to lots of replacement.
Your comments highlight my questions very well:
1. What is the evidence that basal R1a was pretty common in the Yamna region and timescale?
2. What is the explanation for the heavier concentration of basal R1a (and indeed basal R1b) samples in South West Asia (a region in which survival would seem less likely, due to competition from a greater number of other haplogroups)?
3. If basal lines in Ukraine were replaced by surviving R lines, isn't it more likely that the replacers were the only basal R lines that were successful enough to demonstrate survival in some form (i.e. the current Middle Eastern ones)? Especially as the trail from M420 through M459, M198, M417 and Z264 to Z283 is all observable in current day samples of apparently Middle Eastern origin. If you then follow the trails from M417, Z264 and Z283 further, they seem to branch out quickly and directly to North West Europe, the Baltic coast and India - not Ukraine.

epp
02-09-2017, 11:49 PM
Your statements are false. Take a look at the first page of the R1a Project spreadsheet (https://www.familytreedna.com/public/R1a?iframe=yresults).
M420+ M459- : British Isles, Italy
M459+ M198- : Poland, Belarus, Czech Republic, Italy, Russia
M198+ M417- : Germany, Netherlands, Puerto Rico, British Isles, Austria, Macedonia
There's no advantage in getting emotional or argumentative - it can make people ignore or miss what they might not want to see.
You've listed the European countries at the start of the R1a spreadsheet, but overlooked all the Asian ones - Iraq, Qatar, Yemen, Turkey, Lebanon, also Tunisia.
It is also easy to miss that the samples that you do list appear likely to be of predominantly Jewish origin - they are names common to British Sephardis; and Dutch, German and Belarussian Ashkenazis.
Note there's no Ukraine in your list; and the first branches from these basal groups seem to go straight to Scandinavia and the Baltic, with no early groups in the Steppes.

Coldmountains
02-09-2017, 11:58 PM
Yes, sorry - M417, not M459.


Your comments highlight my questions very well:
1. What is the evidence that basal R1a was pretty common in the Yamna region and timescale?
2. What is the explanation for the heavier concentration of basal R1a (and indeed basal R1b) samples in South West Asia (a region in which survival would seem less likely, due to competition from a greater number of other haplogroups)?
3. If basal lines in Ukraine were replaced by surviving R lines, isn't it more likely that the replacers were the only basal R lines that were successful enough to demonstrate survival in some form (i.e. the current Middle Eastern ones)? Especially as the trail from M420 through M459, M198, M417 and Z264 to Z283 is all observable in current day samples of apparently Middle Eastern origin. If you then follow the trails from M417, Z264 and Z283 further, they seem to branch out quickly and directly to North West Europe, the Baltic coast and India - not Ukraine.

Not sure what are you trying to argue here. Ukraine is one of the areas in Europe lowest in R1b so was Yamnaya not R1b? Modern distribution of y-dna is irrelevant and only ancient dna matters. The basal lineages of R1a im SW Asia are interesting but irrelevant they could be very recent and most likely not even predate the Bronze Age. Yamnya was not the only and first steppe culture. I doubt they had much R1a and maybe not even much R1b-L51 . It seems that both were spread by other steppe cultures predating Yamnaya but genetically very similar.

Gravetto-Danubian
02-10-2017, 12:16 AM
It might be worth recalling, before too much time and effort is spent, Epp is prone to odd ideas, and will argue as nauseum despite his lack of familiarity with a subject matter

http://www.anthrogenica.com/showthread.php?6293-Did-haplogroup-I-only-develop-after-the-Ice-Age&p=135545#post135545

Agamemnon
02-10-2017, 12:22 AM
The more common view of experts in this forum is that YFull (slightly) underestimates TMRCAs.

Totally second that, this is quite obvious with J1 for example because we have a set of very reliable TMRCA estimates for each major branch, these estimates are usually underestimated by a margin of ~10% to ~15% on YFull (this goes down with more recent lineages and up with more ancient ones).

lgmayka
02-10-2017, 02:51 AM
There's no advantage in getting emotional or argumentative
I was not being "argumentative," much less "emotional." I stated a simple fact: That your assertions were false. You made the common mistake of using the word all (twice!) without very carefully checking the available data. All means 100%. If you meant most or simply many, you should have written exactly that.


It is also easy to miss that the samples that you do list appear likely to be of predominantly Jewish origin - they are names common to British Sephardis; and Dutch, German and Belarussian Ashkenazis.
That's certainly a very fanciful hypothesis. Do you have evidence that the "predominant" majority of these men are Jewish, besides your personal opinion of their surnames? Did you notice that the given names of their ancestors are
Matthew
Joel
Hugh
Robert
John
Thomas
Bazyl
Grigorij
Garrett
Jan
Lambertus
Joseph
Ivan

In contrast, the given names in the well-known Ashkenazi R-Y2630 clade are:
Wolf
Jehuda
David
Yonkel
Simon
Samuel
Shalom
etc.

Notice the difference.

lgmayka
02-10-2017, 02:38 PM
By the way, I do think that the Southwest Asian examples of R1axM417 deserve further investigation, especially via Big Y testing. We have a few examples, but not nearly enough to come to firm conclusions. Here are the R1axM417 clades on YFull's haplotree:
R-YP4141 (https://yfull.com/tree/R-YP4141/) (subclade of M420): 5 entries, TMRCA 12,600 years
R-YP1272 (https://yfull.com/tree/R-YP1272/) (subclade of M459): 3 entries, TMRCA 7300 years
R-YP1051 (https://yfull.com/tree/R-YP1051/) (subclade of M198): 2 entries, TMRCA 2600 years

epp
02-10-2017, 07:06 PM
I was not being "argumentative," much less "emotional." I stated a simple fact: That your assertions were false. You made the common mistake of using the word all (twice!) without very carefully checking the available data. All means 100%. If you meant most or simply many, you should have written exactly that.
Stating that my assertions were false (in bold) just felt like you were angrily bludgeoning me with your opinion. Perhaps misread you. I am just informally chatting, so forgive me if I use generalisations.


That's certainly a very fanciful hypothesis.
The thing is, I don't fancy this hypothesis. It's not what I expected when I started to research the matter. I'm just going where the data is taking me.


Do you have evidence that the "predominant" majority of these men are Jewish, besides your personal opinion of their surnames? Did you notice that the given names of their ancestors are
Matthew
Joel
Hugh
Robert
John
Thomas
Bazyl
Grigorij
Garrett
Jan
Lambertus
Joseph
Ivan

In contrast, the given names in the well-known Ashkenazi R-Y2630 clade are:
Wolf
Jehuda
David
Yonkel
Simon
Samuel
Shalom
etc.

Notice the difference.
I have to tell you, Igmayka, that not all people of Jewish origin have names like Jehuda, Yonkel and Shalom! I'm also a bit surprised that you don't think names like Matthew, Joel and Joseph would be Jewish. But in any case, the Jewish diasporas occurred 2,000-3,000 years ago, and many of the descendants of these Jews would have mixed in with local populations and not even identified themselves as Jewish at all. That said, I would be surprised if FTDNA's confirmed R1a basal samples Tunisian Mehrez, Belarussian Shpakovsky, Dutch Van Der Water and Scottish Forbes were not of fairly recent Jewish ancestry - the surnames are certainly identifiable as such, and these samples also appear to have a relatively recent MRCA. What else is more likely to have linked these recently-related samples? As I've previously suggested, the narrow link to most European R1a (basal R1a1b1*'s Feist and Myers) also look Jewish, rather than from the Russian Steppes.

I'm not trying to convince anybody of anything. I'm just looking for further evidence/information - although, it's a promising development to see at least one or two of you acknowledging the existence of basal Middle Eastern R1a and keeping an open mind.

epp
02-10-2017, 07:16 PM
It might be worth recalling, before too much time and effort is spent, Epp is prone to odd ideas, and will argue as nauseum despite his lack of familiarity with a subject matter

http://www.anthrogenica.com/showthread.php?6293-Did-haplogroup-I-only-develop-after-the-Ice-Age&p=135545#post135545

Quite amusing to see someone with 2,761 posts complaining that I argue ad nauseam! I hope all those posts have been worth the time and the effort, G-D!

DMXX
02-10-2017, 07:23 PM
I'm not trying to convince anybody of anything. I'm just looking for further evidence/information - although, it's a promising development to see at least one or two of you acknowledging the existence of basal Middle Eastern R1a and keeping an open mind.

The core of the disagreement lies with the consideration of modern parahaplogroup R1a haplotypes from West/Southwest Asia as a counter-weight to the aDNA we have thus far with respect to its' origins.

The holding of said haplotypes as a counter-weight is fallible for two reasons:
1) A lot (but not all as lgmayka points out!) of the basal from West/Southwest Asian R1a haplotypes appear quite drifted. This proved to be the case with R1a*-M420 (picked up by Underhill et al. 2009). All of the samples detected in that paper are from West and Southwest Asia, but the STRs demonstrate they share a relatively recent MRCA in contradistinction with the automatic reference one would hold of this profoundly upstream SNP.
2) None of the aDNA from anywhere in West or Southwest Asia thus far (Israel, Turkey, Armenia, Iran) has produced any Y-DNA R1a subclades. However, all of the pre-EMBA Y-DNA R1a we currently have is located in the swathe of territory that encompasses Eastern Europe.

It is on the basis of the above that many enthusiasts infer that Y-DNA R1a will inevitably be found in Yamnaya. Please note that only the eastern half of the Yamnaya horizon has been sampled to date.

Not impossible that the above will turn out to be a false premise, but as stated recently in another thread, the premise of a prehistoric West or Southwest Asian origin for Y-DNA R1a and R1b continues to become increasingly improbable with every paper lending further credence to reason #2 stated above.

And:

[ADMIN] Let's refrain from personalising this discussion and keeping it subject oriented please, ladies and gents.

epp
02-10-2017, 09:35 PM
The core of the disagreement lies with the consideration of modern parahaplogroup R1a haplotypes from West/Southwest Asia as a counter-weight to the aDNA we have thus far with respect to its' origins.

The holding of said haplotypes as a counter-weight is fallible for two reasons:
1) A lot (but not all as lgmayka points out!) of the basal from West/Southwest Asian R1a haplotypes appear quite drifted. This proved to be the case with R1a*-M420 (picked up by Underhill et al. 2009). All of the samples detected in that paper are from West and Southwest Asia, but the STRs demonstrate they share a relatively recent MRCA in contradistinction with the automatic reference one would hold of this profoundly upstream SNP.
2) None of the aDNA from anywhere in West or Southwest Asia thus far (Israel, Turkey, Armenia, Iran) has produced any Y-DNA R1a subclades. However, all of the pre-EMBA Y-DNA R1a we currently have is located in the swathe of territory that encompasses Eastern Europe.

It is on the basis of the above that many enthusiasts infer that Y-DNA R1a will inevitably be found in Yamnaya. Please note that only the eastern half of the Yamnaya horizon has been sampled to date.

Not impossible that the above will turn out to be a false premise, but as stated recently in another thread, the premise of a prehistoric West or Southwest Asian origin for Y-DNA R1a and R1b continues to become increasingly improbable with every paper lending further credence to reason #2 stated above.

And:

[ADMIN] Let's refrain from personalising this discussion and keeping it subject oriented please, ladies and gents.


My research has used only confirmed, reliable, standardised DNA results, compiled in substantial detail, and released for public consumption in industrial quantities without any apparent agenda (FTDNA databases fit the bill perfectly). This is a summary of its estimates for surviving R1a origins, using ISOGG classifications:
R1/R1a - Western Arabia/Levant (both R1/R1a* and R1a)
R1a - Levant (both R1a* and R1a1)
R1a1 - Levant (both R1a1* and R1a1a)
R1a1a - Levant (both R1a1a* and R1a1a1)
R1a1a1 - Levant (estimated origin point of both its parent R1a1a* & its oldest surviving subclade R1a1a1b)
R1a1a1b - Levant (estimated origin point of both its parent R1a1a1b & its subclade R1a1a1b1*)
R1a1a1b1 - Levant (estimated origin point of its parent R1a1a1b, and closest to its brother subclade R1a1a1b2's estimated point of origin)
R1a1a1b2 - Turkey (point of its greatest diversity and widest range of subclades)
R1a1a1a - North Sea (point of its greatest diversity, subsequently spreading Eastwards into Scandinavia)
R1a1a1b1a - Poland/Baltic (point of its greatest diversity and widest range of subclades, subsequently spreading Eastwards into Northern Russia)

The most obvious and direct route for R1a into Europe and Corded Ware would have been from Turkey and up the Danube into Germany. The large volume of data that I have looked at shows no sign of R1a in Ukraine/South Russia until a few subclades of subclades of R1a1a1b1a arrived there at a much later point.

Perhaps there were some basal R1 & R1a branches in Ukraine/South Russia during Yamna or before then, but the large volume of data that I've worked with shows no estimated sign of its descendants having survived.

This same method of analysis does however estimate R1b-Z2103 to be present in Ukraine/South Russia at the time of Yamna, matching the results of the ancient DNA samples. What I am interested in is to find if there is any reliable data that I have not yet seen that would cause the results of my analysis to be amended.

Michał
02-10-2017, 11:55 PM
My research has used only confirmed, reliable, standardised DNA results, compiled in substantial detail, and released for public consumption in industrial quantities without any apparent agenda (FTDNA databases fit the bill perfectly). This is a summary of its estimates for surviving R1a origins, using ISOGG classifications
I have a strong feeling that your "research" is biased and your reasoning is mostly circular, so I hope you will answer my below question to prove me wrong in this respect.




R1a - Levant (both R1a* and R1a1)
R1a1 - Levant (both R1a1* and R1a1a)
You apparently missed the fact that R1a* (according to ISOGG, which is in fact R1a2) is found in both Europe and Middle East.
Also, could you please show us a list of all those R1a1* (or actually R1a1b) FTDNA kits that were found in the Levant? I hope you don't count Europe nor Northern Africa as Levant.
The Levant origin of R1a1a can also be questioned, which is what I do below.




R1a1a - Levant (both R1a1a* and R1a1a1)
Please let us see your list of those confirmed Levantine R1a1a* (R1a1a2) cases found among the FTDNA customers.
The Levant origin of R1a1a1 can also be questioned, which is what I do below.



R1a1a1 - Levant (estimated origin point of both its parent R1a1a* & its oldest surviving subclade R1a1a1b)
Firstly, you have been so far unable to demonstrate the Levant origin of R1a1a* (for which you need to answer my above questions).
Secondly, the Levant origin of R1a1a1b can also be questioned, which is what I do below.



R1a1a1b - Levant (estimated origin point of both its parent R1a1a1b & its subclade R1a1a1b1*)
Firstly, it is not clear whether you meant "parent R1a1a1b" or rather "parent R1a1a1".
Secondly, in case you really meant "parent R1a1a1b" this simply doesn't make any sense, as your line of reasoning seems to be that R1a1a1b is from Levant because R1a1a1b is from Levant.
Thirdly, in case you meant "parent R1a1a1", you have been (so far) unable to prove that R1a1a1 originated in the Levant (for which you need to answer my above questions).
Fourthly, in this case you totally ignore the European distribution/origin of R1a1a1b's only brother clade (R1a1a1a).
Fifthly, the Levant origin of R1a1a1b1* (I guess you meant R1a1a1b1) can also be questioned, which is what I do below.



R1a1a1b1 - Levant (estimated origin point of its parent R1a1a1b, and closest to its brother subclade R1a1a1b2's estimated point of origin)
Firstly, you have been (so far) unable to prove that the parental clade R1a1a1b originated in the Levant (for which you need to answer my above questions).
Secondly, you totally ignore the predominantly European distribution of R1a1a1b1 itself.
Thirdly, the Levant origin of R1a1a1b2 can also be questioned, which is what I do below.


R1a1a1b2 - Turkey (point of its greatest diversity and widest range of subclades)
You totally ignore the predominantly European distribution of both the brother clade R1a1a1b1 and the uncle clade R1a1a1a.
Also, you don't seem to take into account the regions where the largest number of R1a1a1b2* lineages has been found so far. Please read these two posts in another thread:
http://www.anthrogenica.com/showthread.php?4211-Russian-and-Polish-R1a-Z93*&p=76524&viewfull=1#post76524
http://www.anthrogenica.com/showthread.php?4211-Russian-and-Polish-R1a-Z93*&p=82028&viewfull=1#post82028



The most obvious and direct route for R1a into Europe and Corded Ware would have been from Turkey and up the Danube into Germany.
As has been already mentioned by others, what you suggest is definitely not supported by the ancient DNA data, as R1a is missing in both ancient Turkey and in the Balkans, while being relatively frequent in ancient Eastern Europe.



The large volume of data that I have looked at shows no sign of R1a in Ukraine/South Russia until a few subclades of subclades of R1a1a1b1a arrived there at a much later point.
Firstly, we don't have much data for Ukraine/South Russia until the Yamna period.
Secondly, we still have much more ancient R1a for Eastern Europe than we do for West Asia (where no R1a was found to date).
Thirdly, you seem to be focused exclusively on the Middle East and Ukraine/South Russia, while ignoring the data we have for more northerly regions in Eastern Europe (including Karelia, the Samara region and the Smolensk region), not to mention the Altai region where ancient R1a was also found.


Perhaps there were some basal R1 & R1a branches in Ukraine/South Russia during Yamna or before then, but the large volume of data that I've worked with shows no estimated sign of its descendants having survived.
I guess this is in sharp contrast to the putative ancient Levantine R1a population that has allegedly produced multiple waves of different R1a subclades migrating to Europe and Siberia but has somehow managed to remain undetected by the aDNA studies. ;)

Having said all above, I should also make it clear that you might be right about the early presence of R1a in the Middle East, as we don't have enough data to definitely rule it out. However, it is also quite obvious that what we have right now does not allow us to make any strong conclusions about the very early migrations of the R1a people. So it bothers me a lot when people ignore the data that don't fit their hypothesis, like you do in the above analysis.

epp
02-11-2017, 12:01 PM
Michal, when you say that I am "unable to prove" things, what people on this forum are perhaps missing is that I am not trying to prove anything. There is no assertion that I want to prove. If I had a hypothesis, it was that early development of surviving R1a occurred in Southern Russia and Ukraine within Yamna. However, the FTDNA data that I've analysed does not seem to support this. I am wholly indifferent between one theory and another, and merely interested in what the data suggests. The purpose of this thread is find out if there is anything I have missed - not to engage in squabbles or to submit what people misinterpret as my hypotheses to aggressive questioning.

I think it would make it quite muddly and diverting to respond to each of your list of points one after the other, so I will just give some of the pointers that you request. I'm afraid I can't list every sample I've analysed, which would take too long. By the way, you can find all the samples I used on the FTDNA database - anyone can conduct the same analysis as I have from it - I analysed the concentrations of maximum diversity within unique subclades, working only with the confirmed SNPs and the ISOGG classifications, and brought into my analysis every single sample published (ignoring none). I did not use ancient or other piecemeal sources of DNA data, which seem to be of mixed quality, public availability and reliability.

The term 'basal' can be a bit misleading, but I will use it to identify samples outside the largest, relatively-recent subclades of R1a1a1a, R1a1a1b1a and R1a1a1b2 - Bazqamah, Forbes, Haddad, Abdulla, Travers, Jones, Mehrez, Shpakovsky, Baev, Myllward, Van Der Water, Mihaylov, Wortman, Williams, Feist, Myers, Ledwill, Benbrook.
As you can see even without any quantitative analysis, the range of basal samples doesn't look very East European - indeed, the only three East European samples seem likely to be of Jewish origin, as in fact do the West European samples with similar STR profiles (and also other indeterminate basal samples like Cates).

I can see roughly when the data estimates that surviving R1a arrived in North Western/North Central Europe - R1a1a1 (pre-R1a1a1a) & R1a1a1b1 (pre-R1a1a1b1a) are calculated to have arrived at the North Sea and the Baltic respectively at approximately the same time (Corded Ware). My particular interest is in the significant apparent gaps, and whether these have anything to do with Russia or Yamna. The data estimates a couple of gaps or bottlenecks - R1a1a1b* and R1a1a1b2*, the latter of which seems to have been a slow developer before suddenly flourishing in Southern Asia. In particular, I am wondering if there is any data to indicate that R1a1a1b2 might have originated in the North and was chased Southwards and almost exterminated (perhaps by R1b-Z2103 or by R1a1a1b1a), before later expanding across Asia?

Michał
02-11-2017, 01:08 PM
I'm afraid I can't list every sample I've analysed, which would take too long.
If this is indeed too much work for you, then please show me just some individual examples of the FTDNA customers from the Levant who are securely classified as R1a1* or R1a1a*. I can show you such examples for Europe (including the Eastern Europe), so in case you fail to do the same, this will be a perfect example of how your analysis is biased.


By the way, you can find all the samples I used on the FTDNA database - anyone can conduct the same analysis as I have from it
I would be definitely able to find all such cases if only they existed. The problem is there are no such cases in FTDNA so far, while there are quite many examples of R1a* (R1a2), R1a1* (R1a1b) or R1a1a* (R1a1a2) for Europe, so although I admit your hypothesis has not been disproved yet, it is quite obvious that your analysis was strongly biased.

Generalissimo
02-11-2017, 01:09 PM
I can see roughly when the data estimates that surviving R1a arrived in North Western/North Central Europe - R1a1a1 (pre-R1a1a1a) & R1a1a1b1 (pre-R1a1a1b1a) are calculated to have arrived at the North Sea and the Baltic respectively at approximately the same time (Corded Ware). My particular interest is in the significant apparent gaps, and whether these have anything to do with Russia or Yamna. The data estimates a couple of gaps or bottlenecks - R1a1a1b* and R1a1a1b2*, the latter of which seems to have been a slow developer before suddenly flourishing in Southern Asia. In particular, I am wondering if there is any data to indicate that R1a1a1b2 might have originated in the North and was chased Southwards and almost exterminated (perhaps by R1b-Z2103 or by R1a1a1b1a), before later expanding across Asia?

No, the most common R1a lineages around today expanded from the steppe with Corded Ware, which was a population very closely related to Yamnaya, or perhaps even an Yamnaya offshoot.

The earliest Corded Ware genomes are almost identical to Yamnaya.

epp
02-11-2017, 01:45 PM
No, the most common R1a lineages around today expanded from the steppe with Corded Ware, which was a population very closely related to Yamnaya, or perhaps even an Yamnaya offshoot.

The earliest Corded Ware genomes are almost identical to Yamnaya.
I haven't said anything about Corded Ware genomes. All I noted is that my calculations from FTDNA data estimate that R1a arrived at the North Sea and the Baltic at approximately the same time as the Corded Ware period.

Generalissimo
02-11-2017, 01:52 PM
I haven't said anything about Corded Ware genomes.

Yes, but I did. And it's something you should consider.

epp
02-11-2017, 02:27 PM
If this is indeed too much work for you, then please show me just some individual examples of the FTDNA customers from the Levant who are securely classified as R1a1* or R1a1a*. I can show you such examples for Europe (including the Eastern Europe), so in case you fail to do the same, this will be a perfect example of how your analysis is biased.

I would be definitely able to find all such cases if only they existed. The problem is there are no such cases in FTDNA so far, while there are quite many examples of R1a* (R1a2), R1a1* (R1a1b) or R1a1a* (R1a1a2) for Europe, so although I admit your hypothesis has not been disproved yet, it is quite obvious that your analysis was strongly biased.
Do you really imagine I'd be so biased that I would make up a lot of random names?
As these samples apparently don't exist, perhaps I am just imagining seeing them on https://www.familytreedna.com/public/R1a?iframe=yresults -
Bazqamah M11251, Forbes 219391, Haddad 310205, Abdulla N114240, Travers B10992, Jones B13606, Mehrez E18637, Shpakovsky N86494 ... do I really have to go on?
A few of the samples are unnamed on the FTDNA spreadsheet, but the names are easily obtainable elsewhere on the internet.

If you have STRs and confirmed SNPs for these basal samples of R1a, then yes I would be interested to see them. As I've suggested, I would be even more interested to see samples of R1a1a1b* or outlying samples of R1a1a1b2*, if there are any that I've missed - whether European or otherwise! I must admit, though, that chatting on this forum is increasing my scepticism about miscellaneous bits of DNA data posted on the internet, as geneticists all seem hung up on their own prior conclusions and become shrouded in a grey mist whenever any evidence is noted that it inconsistent with them.

epp
02-11-2017, 02:41 PM
Yes, but I did. And it's something you should consider.
Perhaps you're right. Do you know what the genomic similarities are, and do they relate to y-dna?
If Yamna R1b are nearly genomically identical to Corded Ware R1a, then presumably the close similarities are not going to have been passed on through y-dna.

Generalissimo
02-11-2017, 03:32 PM
If Yamna R1b are nearly genomically identical to Corded Ware R1a, then presumably the close similarities are not going to have been passed on through y-dna.

Genome-wide ancestry is not passed on through Y-DNA. The two aren't linked, although they can correlate. And uniparental markers, especially Y-haplogroups, are susceptible to severe founder effects, which often happen during expansions. In other words, there was probably both R1a and R1b in both Yamnaya and Corded Ware, but the samples we have represent clans affected by paternal founder effects.

epp
02-11-2017, 04:08 PM
Genome-wide ancestry is not passed on through Y-DNA. The two aren't linked, although they can correlate. And uniparental markers, especially Y-haplogroups, are susceptible to severe founder effects, which often happen during expansions. In other words, there was probably both R1a and R1b in both Yamnaya and Corded Ware, but the samples we have represent clans affected by paternal founder effects.
Possibly. Although R1a and R1b don't seem to have developed in tandem - they separated around 15,000 years ago and have since migrated to different parts of the world, rarely together. Of course, there is likely to have been contact, but I can't see much DNA evidence that they merged together substantially until fairly recently in Scandinavia & Germany.

Michał
02-11-2017, 04:57 PM
Do you really imagine I'd be so biased that I would make up a lot of random names?
I don't think you made up any names. You just failed to show us any FTDNA customers from the Levant who would be classified as R1a1* or R1a1a* (which is what you have been asked to do).


As these samples apparently don't exist, perhaps I am just imagining seeing them on https://www.familytreedna.com/public/R1a?iframe=yresults -
Bazqamah M11251, Forbes 219391, Haddad 310205, Abdulla N114240, Travers B10992, Jones B13606, Mehrez E18637, Shpakovsky N86494

First, let me remind you that I have asked you specifically to show any examples of R1a1* or R1a1a* from the Levant, so what you show above does not meet these criteria. The first six people from your set (ie. Bazqamah M11251, Forbes 219391, Haddad 310205, Abdulla N114240, Travers B10992, Jones B13606) are all classifies as R1a*/R1a2, so they do not fall neither under R1a1* nor under R1a1a*. Also, three lineages from that group (ie. Forbes 219391, Travers B10992 and Jones B13606) are from Western Europe and don't show any Middle Eastern/Levant ancestry.

The two other FTDNA customers mentioned above (Mehrez E18637 and Shpakovsky N86494) are indeed both confirmed as R1a1*/R1a1b, but neither of them is from the Levant. Instead, Mehrez is from Northern Africa (Tunisia) and Shpakovsky is from Eastern Europe (Poland/Belarus).



... do I really have to go on?
Please do, unless you admit that you don't know any FTDNA customers from the Levant who would be classified as R1a1* or R1a1a*.



If you have STRs and confirmed SNPs for these basal samples of R1a, then yes I would be interested to see them.
I am one of the admins of that R1a project at FTDNA from which your data come. Thus, I can assure you that we have 7 cases of R1a1*/R1a1b and 13 cases of R1a1a*/R1a1a2, and neither of them is from the Levant (although we have two R1a1*/R1a1b cases from North Africa, in addition to 5 customers from Europe). All 13 R1a1a*/R1a1a2 lineages in our project are from Europe (mostly from Western Europe).

I guess you will be interested in some extra information about the rare R1a cases found in the Middle East, so let me mention that I am actually aware of one R1a1*/R1a1b case from Iraq, but this is a research sample, not an FTDNA customer. Similarly, I know about two potential R1a1*/R1a1b research samples from Cyprus (where a large population was tested with STRs). These two Cypriot lineages seemed to be very closely related to each other, as indicated by their almost identical STR haplotypes.



As I've suggested, I would be even more interested to see samples of R1a1a1b* or outlying samples of R1a1a1b2*, if there are any that I've missed - whether European or otherwise!
There are no R1a1a1b* (Z645*) cases known to us, neither in the literature nor in the FTDNA projects. All R1a1a1b (Z645) members tested so far are either R1a1a1b1 (Z283+) or R1a1a1b2 (Z93+). As for R1a1a1b2* (Z93*), please read those two posts I have already recommended to you.


I must admit, though, that chatting on this forum is increasing my scepticism about miscellaneous bits of DNA data posted on the internet, as geneticists all seem hung up on their own prior conclusions and become shrouded in a grey mist whenever any evidence is noted that it inconsistent with them.
I would rather strongly disagree with your above opinion. All forumers who responded to your first post are much more knowledgeable about R1a than you are, yet they have decided to share their knowledge with you without "hiding" any data that could have spoken in favor of your hypothesis. By contrast, what you have shown to us was a result of an extremely biased analysis. And, more importantly, you seem to be unable to admit that you have been wrong about it.

epp
02-11-2017, 06:32 PM
I don't think you made up any names. You just failed to show us any FTDNA customers from the Levant who would be classified as R1a1* or R1a1a* (which is what you have been asked to do).
I tried to be helpful and listed my samples that I said fell outside of R1a1a1a, R1a1a1b1a and R1a1a1b2 (Bazqamah, Forbes, Haddad, Abdulla, Travers, Jones, Mehrez, Shpakovsky, Baev, Myllward, Van Der Water, Mihaylov, Wortman, Williams, Feist, Myers, Ledwill, Benbrook) and advised "you can find all the samples I used on the FTDNA database". You then replied "I would be definitely able to find all such cases if only they existed" They do exist, and the majority of them are R1a1* or R1a1a*, e.g. Mehrez is M459+ and M512-, as you have confirmed below.


Also, three lineages from that group (ie. Forbes 219391, Travers B10992 and Jones B13606) are from Western Europe and don't show any Middle Eastern/Levant ancestry.
The two other FTDNA customers mentioned above (Mehrez E18637 and Shpakovsky N86494) are indeed both confirmed as R1a1*/R1a1b, but neither of them is from the Levant. Instead, Mehrez is from Northern Africa (Tunisia) and Shpakovsky is from Eastern Europe (Poland/Belarus).
One of FTDNA's samples is Shalom ben Gershon Levie-Grin - as he's from Holland, does that mean to you that his DNA is not of Jewish/Levantine origin? The names Mehrez and Shpakovsky too are both indicative of likely Jewish origin. And the other names are also commonly held by Jews - in particular, Forbes, which is of Sephardi origin.


I guess you will be interested in some extra information about the rare R1a cases found in the Middle East, so let me mention that I am actually aware of one R1a1*/R1a1b case from Iraq, but this is a research sample, not an FTDNA customer. Similarly, I know about two potential R1a1*/R1a1b research samples from Cyprus (where a large population was tested with STRs). These two Cypriot lineages seemed to be very closely related to each other, as indicated by their almost identical STR haplotypes.
This might be useful, if you were able to give me the full SNP and STR profiles.


There are no R1a1a1b* (Z645*) cases known to us, neither in the literature nor in the FTDNA projects. All R1a1a1b (Z645) members tested so far are either R1a1a1b1 (Z283+) or R1a1a1b2 (Z93+).
Thank you for confirming this.


As for R1a1a1b2* (Z93*), please read those two posts I have already recommended to you.
If these samples are all on the main R1a FTDNA database, then my calculations will have already included them. If you think there are any particular outliers or significant samples I might have missed that would indicate a Northern or European origin for this subclade, I would be very interested to be advised of which ones.


All forumers who responded to your first post are much more knowledgeable about R1a than you are, yet they have decided to share their knowledge with you without "hiding" any data that could have spoken in favor of your hypothesis.
It would have been a more pleasant experience if some of them could have shared what they had to say in a nicer and more respectful way. And I have to say that acknowledgement that basal R1a subclades existed at all in Asia seemed rather grudging and slow to come by.


By contrast, what you have shown to us was a result of an extremely biased analysis. And, more importantly, you seem to be unable to admit that you have been wrong about it.
This is a prime example. When people simply ask questions and share the results of their analysis without expressing any conclusions, only to be told that all other people on the thread are much more knowledgeable than them, that they are unable to admit that they have been wrong and that their analysis is extremely biased, it is not particularly conducive to encouraging them to engage further on the forum.

Michał
02-11-2017, 08:36 PM
I tried to be helpful and listed my samples that I said fell outside of R1a1a1a, R1a1a1b1a and R1a1a1b2
Please read my previous posts again. I have specifically asked yo to list the R1a1* and R1a1a* members from the Levant (and I have repeated this on several occasions, so you would not have any excuse when not complying to this). Thus, when your answer was about some R1a* members or the R1a1*/R1a11b members from Europe and North Africa (but not from the Levant!), this was not what you were supposed to show us. Also, you haven't mentioned at that time that these are just people from very different regions (including Europe and Africa) who "fell outside of R1a1a1a, R1a1a1b1a and R1a1a1b2", as this was not what you were supposed to include in your list.


(Bazqamah, Forbes, Haddad, Abdulla, Travers, Jones, Mehrez, Shpakovsky, Baev, Myllward, Van Der Water, Mihaylov, Wortman, Williams, Feist, Myers, Ledwill, Benbrook) and advised "you can find all the samples I used on the FTDNA database". You then replied "I would be definitely able to find all such cases if only they existed"
This is simply not true. The above reply of mine was posted before you have provided any names, and what I wrote was specificaly related to the R1a1* and R1a1a* FTDNA customers from the Levant and not to the names from your list that included people either from other clades/paragroups or from other regions.



They do exist, and the majority of them are R1a1* or R1a1a*
It is obvious that such R1a1* or R1a1a* FTDNA customers do exist, and there are 20 of them in our R1a project (as I have explained in my previous post), but neither of them is from the Levant, contrary to what you have claimed.



e.g. Mehrez is M459+ and M512-, as you have confirmed below.
Yes, but he is from North Africa (Tunisia), not from the Levant (and you were supposed to show us such lineages found in the Levant).



One of FTDNA's samples is Shalom ben Gershon Levie-Grin - as he's from Holland, does that mean to you that his DNA is not of Jewish/Levantine origin?
If your list included any R1a1* or R1a1a* members like him, I would most likely count him as of Middle Eastern origin, However, there were no such R1a1* or R1a1a* people included in your list, so all this is simply irrelevant.



The names Mehrez and Shpakovsky too are both indicative of likely Jewish origin.
This is incorrect. Mehrez is definitely an Arab surname that is very common in North Africa (and especially in Tunisia). Shpakovsky/Szpakowski is a typical (ie. quite common) Polish surname that has nothing to do with Jewish ancestry. The spelling was changed to Shpakovsky only after his ancestors moved to America.



And the other names are also commonly held by Jews - in particular, Forbes, which is of Sephardi origin.
This is again incorrect. Forbes is a surname of Scottish/Gaellic origin:
http://www.surnamedb.com/Surname/Forbes
https://en.wikipedia.org/wiki/Forbes_(name)
https://en.wikipedia.org/wiki/Clan_Forbes

Claiming that all European or African R1a1*/R1a1b and of R1a1a*/R1a1a2 lineages are ("by definition") of Jewish ancestry is like claiming that all Levantine R1a cases descend from the European crusaders.



This might be useful, if you were able to give me the full SNP and STR profiles.
Sure, I will try to dig it out for you when finding a free minute.



If these samples are all on the main R1a FTDNA database, then my calculations will have already included them.
Not necessarily. There are many project members whose personal settings make their results visible only to other project members.



If you think there are any particular outliers or significant samples I might have missed that would indicate a Northern or European origin for this subclade, I would be very interested to be advised of which ones.
I will try to update my old list of all known Z93* and Z93(xZ94) lineages when finding some free time.

epp
02-11-2017, 10:04 PM
Please read my previous posts again. I have specifically asked yo to list the R1a1* and R1a1a* members from the Levant (and I have repeated this on several occasions, so you would not have any excuse when not complying to this). Thus, when your answer was about some R1a* members or the R1a1*/R1a11b members from Europe and North Africa (but not from the Levant!), this was not what you were supposed to show us. Also, you haven't mentioned at that time that these are just people from very different regions (including Europe and Africa) who "fell outside of R1a1a1a, R1a1a1b1a and R1a1a1b2", as this was not what you were supposed to include in your list.
Michal, I wonder this is a language thing. In England, when you are conversing with people, and you tell them what they are "supposed to" do and that they have no "excuse" in "not complying", it is seen as quite impolite and dictatorial, but perhaps this is the style of conversation in Poland. As an Englishman, I feel at liberty to say whatever I think I should say, without being limited to having to follow strict instructions.


This is simply not true. The above reply of mine was posted before you have provided any names
At the risk of being pedantic, I provided the names in post 25. The reply of yours was in post 26 - i.e. after I provided the names.


Mehrez is definitely an Arab surname that is very common in North Africa (and especially in Tunisia). Shpakovsky/Szpakowski is a typical (ie. quite common) Polish surname that has nothing to do with Jewish ancestry. The spelling was changed to Shpakovsky only after his ancestors moved to America.
This is again incorrect. Forbes is a surname of Scottish/Gaellic origin:
You could be right, but unless you know the detailed ancestry of these people going back 2,000 years, it is presumptuous to say definitely. Mehrez is also a common Sephardi Jewish surname. I've know of Jews called Shpakovsky. And Forbes is also documented as a name taken on by Jews (a corruption of the Sephardi name Farrabas). I certainly don't know for sure either, but the STRs of the first two are also very similar, and a substantial Jewish community is one of the few things linking Tunisia and Poland/Belarus.


Sure, I will try to dig it out for you when finding a free minute.
Thank you. There's no hurry.


I will try to update my old list of all known Z93* and Z93(xZ94) lineages when finding some free time.
Again, thanks. By the way, I've noticed a few apparent inconsistencies with SNP readings on the FTDNA website which might have skewed my calculations, so I think I'll run the results again excluding these. As you are the site administrator, I'll get back to you on that.

lgmayka
02-11-2017, 11:58 PM
The names Mehrez and Shpakovsky too are both indicative of likely Jewish origin. And the other names are also commonly held by Jews - in particular, Forbes, which is of Sephardi origin.
American culture considers it offensive to aggressively apply one's amateur opinion of surname origins to people one does not know. I note that you do not present a single shred of evidence that these men have Jewish patrilineage, other than your own amateur attribution of their surnames. I must point out that you are effectively discouraging people from posting their DNA results publicly at all, precisely because kibitzers may jump to hasty and unjustified conclusions.

Please consider both cases:
- Let's first say that one of these men really is aware of Jewish patrilineage. He has not joined any specifically Jewish projects, and has in no other way marked himself as (historically) Jewish. Now you come along and attempt to "out" him based solely on your amateur opinion of surnames. I suspect he would be offended.
- Let's now say that most or all of these men are not Jewish, and find no evidence of Jewish patrilineage in their genealogical investigation. You come along and assign them an incorrect religious/ethnic tag, based solely on your amateur attribution of surnames. I suspect they might be downright angry.


When people simply ask questions and share the results of their analysis without expressing any conclusions
As I see it, you jumped to two different conclusions, neither of them justified.
1) You assigned large numbers of men to Jewish patrilineage based solely on amateur surname attribution.
2) You asserted that such Jewish patrilineage must originate in the Levant.

(2) deserves a little more discussion. Let's say (improbably!) that your assertion of at least one of these cryptic medieval Jewish origins is correct. Without specific correlative examples from other Levantine populations, the most logical working hypothesis (by Occam's Razor) would be that this R1a lineage introgressed into the Jewish community from elsewhere within historical times--from Belarus, Spain, Italy, Greece, or even Iraq (during the Babylonian Exile). I would see no reason to assume that such an uncharacteristic lineage has a specifically Levantine origin.

epp
02-12-2017, 01:16 AM
American culture considers it offensive to aggressively apply one's amateur opinion of surname origins to people one does not know. I note that you do not present a single shred of evidence that these men have Jewish patrilineage, other than your own amateur attribution of their surnames. I must point out that you are effectively discouraging people from posting their DNA results publicly at all, precisely because kibitzers may jump to amateur conclusions.

Please consider both cases:
- Let's first say that one of these men really is aware of Jewish patrilineage. He has not joined any specifically Jewish projects, and has in no other way marked himself as (historically) Jewish. Now you come along and attempt to "out" him based solely on your amateur opinion of surnames. I suspect he would be offended.
- Let's now say that most or all of these men are not Jewish, and find no evidence of Jewish patrilineage in their genealogical investigation. You come along and assign them an incorrect ethnic tag, based solely on your amateur attribution of surnames. I suspect they might be downright angry.
I suppose, if they are absurdly anti-semitic, they might be angry at the mere possibility of having 0.0000000000000000001% Jewish ancestry, but don't we pretty much all have at least that - or is that me now unfairly "outing" the majority of the world's population? What is odd is that, to you, my estimated guess that a person might have one Jewish ancestor from 2,000 years ago is aggressively offensive enough to deserve a rebuke, yet another's claim that someone is definitely Arab or "has nothing to do with" Jewish ancestry is not sufficiently offensive. I personally don't see that any enquiries like this are offensive. If people are unhappy with others looking at their DNA ancestry and studying it, why would they post their names, countries of origin and DNA details online?


As I see it, you jumped to two different conclusions, neither of them justifiable.
1) You assigned large numbers of men to Jewish patrilineage based solely on amateur surname attribution.
2) You asserted that such Jewish patrilineage must originate in the Levant.
I don't know how many times I'm going to have to repeat this - I assert nothing. All I am doing is asking questions, based on estimations.


(2) deserves a little more discussion. Let's say (improbably!) that your assertion of one or more of these cryptic medieval Jewish origins is correct. Without specific correlative examples from other Levantine populations, the most logical hypothesis (by Occam's Razor) would be that these R1a lineages introgressed into the Jewish community from elsewhere within historical times--from Belarus, Spain, Italy, Greece, or even Iraq (during the Babylonian Exile).
Yes, any of these are possible, but there are few other samples of these basal subclades found anywhere else to cause me to best-guess that it came from there instead - apart from perhaps Haddad, Abdulla & Ali (also Middle East) or Travers and Jones (seemingly from Britain).

Theramster
02-12-2017, 04:05 AM
Michal,

You said
"
I guess you will be interested in some extra information about the rare R1a cases found in the Middle East, so let me mention that I am actually aware of one R1a1*/R1a1b case from Iraq, but this is a research sample, not an FTDNA customer. Similarly, I know about two potential R1a1*/R1a1b research samples from Cyprus (where a large population was tested with STRs). These two Cypriot lineages seemed to be very closely related to each other, as indicated by their almost identical STR haplotypes"

I understand my relationship to the Cypriots' samples. Would you please illuminate me on my relationship to the Iraqi sample?

I also wish to let Epp know that I don't have any Jewish descent that I'm aware of. Guirguis is Coptic for Georges.

I look forward to hearing from you. Thank you.

lgmayka
02-12-2017, 12:26 PM
I suppose, if they are absurdly anti-semitic
Please don't throw harsh accusations at people who simply want to be treated with respect. No one wants to be misidentified and miscategorized by uninformed outsiders.


another's claim that someone is definitely Arab or "has nothing to do with" Jewish ancestry
That explanation came from a project administrator who has actual knowledge of the cases and has communicated with the individuals.


If people are unhappy with others looking at their DNA ancestry and studying it, why would they post their names, countries of origin and DNA details online?
People who publish their data expect to be treated with respect, not with sweeping naive generalizations claiming that "you people aren't who you say you are."

But at least we can agree upon the need for further investigation of unusual R1a cases appearing in the Middle East and North Africa:
M420xM459 in Iraq, Qatar, Yemen, Turkey, and Lebanon
M459xM198 in Egypt and Tunisia
etc.
They are indeed intriguing.

epp
02-12-2017, 12:59 PM
Michal,

You said
"
I guess you will be interested in some extra information about the rare R1a cases found in the Middle East, so let me mention that I am actually aware of one R1a1*/R1a1b case from Iraq, but this is a research sample, not an FTDNA customer. Similarly, I know about two potential R1a1*/R1a1b research samples from Cyprus (where a large population was tested with STRs). These two Cypriot lineages seemed to be very closely related to each other, as indicated by their almost identical STR haplotypes"

I understand my relationship to the Cypriots' samples. Would you please illuminate me on my relationship to the Iraqi sample?

I also wish to let Epp know that I don't have any Jewish descent that I'm aware of. Guirguis is Coptic for Georges.

I look forward to hearing from you. Thank you.

Thank you for this information. I too would be interested to learn of the degree of similarity between these samples, but of course only if the people concerned are happy for such information to be shared.

epp
02-12-2017, 02:15 PM
Are you now throwing harsh accusations at people who simply want to be treated with respect? No one wants to be misidentified and miscategorized by uninformed kibitzers.

That explanation came from a project administrator who has actual knowledge of the cases and has communicated with the individuals.

People who publish their data expect to be treated with respect, not with sweeping naive generalizations claiming that "you people aren't who you say you are."
I seem to be routinely misunderstood on this forum. The subject of my thread is R1a and the Yamna culture. Other forum users diverted me into a subject I had no intention of discussing - explaining my best guesses about the potential Middle Eastern roots of R1a and identifying the samples I relied on, and since then I have been pilloried for trying to do what I was asked.

No one is able to assuredly identify/categorise the geographical locations and ethnicities of all their ancestors going back to prehistoric times. Although obviously I cannot, nor can project administrators, nor the individuals themselves. But this does not stop anyone from speculating on a most likely scenario, based on the data publicly available, and so potentially gaining a greater understanding of human history.

Let's get back to the subject of this thread. Let's assume that none of my basal samples were of Jewish ancestry. This adjusts the results of my analysis of FTDNA data, estimating the European arrival of surviving R1a to be earlier, and the point of its origin to be North West Europe. This time, I am not going to be diverted into debating this side issue - but, for your information, my estimate is based on the location of the basal samples, the estimated calculated origin point of R1a1a1a, the location of the R1a1a1b1* samples, and the calculated concentration of maximum diversity of R1a1a1b1a. My calculations estimate that the successful groups R1a1a1a and R1a1a1b1a developed Eastwards through the Baltic and into Northern Russia. The significant point is that I have found no data to support the estimated existence of any surviving R1a in Ukraine and Southern Russia during the Yamna period, and I am enquiring to see if there is any such data.

My guess is that the significant mystery is what happened to R1a1a1b (xR1a1a1b1). Michal has confirmed there are no R1a1a1b* samples on the FTDNA spreadsheet, and my calculations estimate that R1a1a1b1a developed substantially before the remaining R1a1a1b (xR1a1a1b1) starting developing from a MRCA of R1a1a1b2.

Michal - As you have confirmed, there is a substantial FTDNA sample size for R1a1a1b2*, making me more confident in the results of my calculations. I noticed a few apparent inconsistencies between the FTDNA-reported results and ISOGG classifications, so I have excluded these samples from my calculations and have run the analysis again with all the remaining FTDNA data. This came out with an almost identical result - a estimated locational origin in Turkey with a relatively recent MRCA. I am interested in how R1a1a1b2 might have arrived there - especially if my previous estimations of Levantine basal origin (based on limited information) are incorrect, and if a distant North West European origin for R1a1a1b were correct.

Please, I am not looking for more dispute about side issues, just for any further data/information on the subject I'm enquiring about.

Michał
02-12-2017, 03:00 PM
Would you please illuminate me on my relationship to the Iraqi sample?
It seems that my memory has failed me and this potential YP1272 STR haplotype was from Iran (not from Iraq). In the paper by Underhill et al. (2014), he has been listed as Iranian Azeri. Importantly, there was also another potential YP1272 STR haplotype, this time of North-West Caucasian (Kabardin) origin. They both were tested M420+ SRY10831.2+ M198-, so either R1a1* or R1a1b/R-YP1272.

As for their relationship to the previously known members of clade YP1272 (including your early diverged YP1272* lineage), I suspect they both belong to a more basal subclade under YP1272 (or even under M459), so they are almost certainly negative for YP1276, while their YP1272 status is hard to predict. They both share three STR innovations that are absent in all members of clade YP1272 from our project, and these are DYS389I/a>16(=17 or 18), DYS389b<17(=13 or 14) and DYS458<16(=15). Although these two lineages seem to belong to a common subclade under M459/YP1272, they are not very closely related to each other. Here are their STR haplotypes:

13 24 15 11 12 12 13 12 17 13 30 15 16 21 10 11 15 11 23
13 25 15 10 12 12 13 12 18 13 32 15 16 ND 10 ND ND 11 23

(DYS393 DYS390 DYS19 DYS391 DYS385a DYS385b DYS388 DYS439 DYS389I DYS392 DYS389II DYS458 DYS437 DYS448 DYS461 YGATAH4 DYS456 DYS438 DYS635)

Apart from the above two STR haplotypes, Underhill identified four more R1a1* individuals (all from Iran but not of Azeri ancestry), so together they found five (5/1765; 0.28%) R1a1* individuals in Iran and one (1/141; 0.7%) R1a1* individual in Kabardia. They also tested quite large populations from other parts of the Middle East, Central Asia, South Asia, Northern Africa and Europe, including for example Turkey (645), Armenia(427) and Egypt (147). Among the tested Jewish people there was a group of 610 non-Ashkenazi Jews (Spanish exiles, North Africa, Near East, Yemen), 119 Mountain Jews (Caucasus) and 76 Indian Jews.



I also wish to let Epp know that I don't have any Jewish descent that I'm aware of. Guirguis is Coptic for Georges.

You are one of those who are not visible to non-members, so Epp could have not been aware of your existence until now. I would recommend changing your personal settings at FTDNA, in case you want other people to see your STR results.

Michał
02-12-2017, 03:51 PM
Mehrez is also a common Sephardi Jewish surname.
I really doubt this surname is as common among the Sephardi Jews as it is common among the Arabs in Tunisia and generally in North Africa (could you please provide any numbers for the Sephardi Jewish community). Since Jews were frequently adopting local surnames in different countries, it is not uncommon for them to use non-Jewish surnames, but this does not make any Arab named Mehrez, any Pole named Szpakowski nor any Briton named Forbes to be more likely a Jew than a man of Arab, Polish or British ancestry, respectively.


I've know of Jews called Shpakovsky.
Again, there is a difference between a Jewish name and a common Polish name that could have been taken over by a Jewish family. There are about 3500 people called Szpakowski in Poland and it seems that this surname is also quite common in Belarus. We have no reasons to doubt that the vast majority of them descend from non-Jewish patrilineages (just like it is the case for the entire Polish or Belarusian population). I would even say that it is actually slightly more likely that the Jew you know has some distant paternal ancestors of Polish/Belarusian origin rather than vice versa.


I certainly don't know for sure either, but the STRs of the first two are also very similar, and a substantial Jewish community is one of the few things linking Tunisia and Poland/Belarus.
These two lineages diverged about 4000 years ago (https://www.yfull.com/tree/R-YP1272/).

Michał
02-12-2017, 04:10 PM
I seem to be routinely misunderstood on this forum.
Have you considered a possibility that it is the other way around, ie. it is you who doesn't seem to understand what we are trying to communicate to you?


I noticed a few apparent inconsistencies between the FTDNA-reported results and ISOGG classifications, so I have excluded these samples from my calculations and have run the analysis again with all the remaining FTDNA data.
Could you please elaborate a bit on this? The ISOGG haplotree is definitely outdated, so excluding any samples only because they don't fit the classification suggested by ISOGG doesn't seem to be a good idea. I would rather recommend using the YFull tree as your reference point (https://www.yfull.com/tree/R-Z93/).


Please, I am not looking for more dispute about side issues, just for any further data/information on the subject I'm enquiring about.
Will you consider a critical evaluation of your approach (including both the methodology you use and and the data you include) a side issue or rather a valuable contribution from our side?

vettor
02-12-2017, 05:44 PM
I suppose, if they are absurdly anti-semitic, they might be angry at the mere possibility of having 0.0000000000000000001% Jewish ancestry, but don't we pretty much all have at least that - or is that me now unfairly "outing" the majority of the world's population? What is odd is that, to you, my estimated guess that a person might have one Jewish ancestor from 2,000 years ago is aggressively offensive enough to deserve a rebuke, yet another's claim that someone is definitely Arab or "has nothing to do with" Jewish ancestry is not sufficiently offensive. I personally don't see that any enquiries like this are offensive. If people are unhappy with others looking at their DNA ancestry and studying it, why would they post their names, countries of origin and DNA details online?


I don't know how many times I'm going to have to repeat this - I assert nothing. All I am doing is asking questions, based on estimations.


Yes, any of these are possible, but there are few other samples of these basal subclades found anywhere else to cause me to best-guess that it came from there instead - apart from perhaps Haddad, Abdulla & Ali (also Middle East) or Travers and Jones (seemingly from Britain).

You are wasting your time with linking DNA with anything that includes nationality, religion, ethnicity or surnames , you only have geographical areas to link your ancient DNA to you or others

epp
02-12-2017, 06:31 PM
I really doubt this surname is as common among the Sephardi Jews as it is common among the Arabs in Tunisia and generally in North Africa (could you please provide any numbers for the Sephardi Jewish community). Since Jews were frequently adopting local surnames in different countries, it is not uncommon for them to use non-Jewish surnames, but this does not make any Arab named Mehrez, any Pole named Szpakowski nor any Briton named Forbes to be more likely a Jew than a man of Arab, Polish or British ancestry, respectively.


Again, there is a difference between a Jewish name and a common Polish name that could have been taken over by a Jewish family. There are about 3500 people called Szpakowski in Poland and it seems that this surname is also quite common in Belarus. We have no reasons to doubt that the vast majority of them descend from non-Jewish patrilineages (just like it is the case for the entire Polish or Belarusian population). I would even say that it is actually slightly more likely that the Jew you know has some distant paternal ancestors of Polish/Belarusian origin rather than vice versa.


These two lineages diverged about 4000 years ago (https://www.yfull.com/tree/R-YP1272/).
You could be right. There's not enough data to come to any confident conclusion, and Tunisia and Belarus don't seem linked in any obvious way to offer an clear explanation for this specific divergence. (My own divergence estimate is 3,200 years, i.e. more recent than the major R1a divergences in Northern Europe.)

epp
02-12-2017, 06:56 PM
Have you considered a possibility that it is the other way around, ie. it is you who doesn't seem to understand what we are trying to communicate to you?
Perhaps it is both.


Could you please elaborate a bit on this? The ISOGG haplotree is definitely outdated, so excluding any samples only because they don't fit the classification suggested by ISOGG doesn't seem to be a good idea. I would rather recommend using the YFull tree as your reference point (https://www.yfull.com/tree/R-Z93/).
I think the queries were (i) samples M7180, 210861 & M6507 (described as Z94 on the spreadsheet, when the SNPs show them as Z94-), and (ii) 297806, 179005 & B2586 with SNPs of Z94-, but positive for some of Z94's subclades (like Z2122 and Z2123), per the ISOGG classification. However, for some reason, my computer won't let me onto the website to check. It is not significant, as excluding these few queries does not make a difference to the outcome.


Will you consider a critical evaluation of your approach (including both the methodology you use and and the data you include) a side issue or rather a valuable contribution from our side?
It depends. If it relates to the subject of the thread and offers a positive suggestion for improvement, then it is potentially a valuable contribution. If it does not really relate to the thread and offers no suggestions or better alternatives, or if it is just to make a negative comment about my character or expertise, then I would consider it a side issue. I'm afraid your expression "from our side" does not fill me with great confidence, because it seems to indicate that you consider us as somehow on opposing sides - me against the rest of you. I am not really interested in argumentative game play.

epp
02-12-2017, 07:05 PM
You are wasting your time with linking DNA with anything that includes nationality, religion, ethnicity or surnames , you only have geographical areas to link your ancient DNA to you or others
It depends on the circumstances. The presence of divergent R1b or E1b1a samples in USA obviously does not indicate North American origin for these haplogroups. And a name like Yonkel, for instance, could suggest ethnicity, regardless of geographical location. As with any investigation, the more data you have to work with, the easier it becomes to arrive at estimates you can be confident about.

Michał
02-12-2017, 10:02 PM
(i) samples M7180, 210861 & M6507 (described as Z94 on the spreadsheet, when the SNPs show them as Z94-),

I guess M7180 should be placed in the below category II. His Z94- result shown on the SNP page is an error. Apparently, there is a bug in the FTDNA software.
As for 210861 & M6507, they have not been tested for Z94 yet (so they are not shown as Z94- on our project's SNP page) and our spreadsheet doesn't show them as Z94.



and (ii) 297806, 179005 & B2586 with SNPs of Z94-, but positive for some of Z94's subclades (like Z2122 and Z2123), per the ISOGG classification.

They are all wrongly shown as Z94- on the SNP page, so this is a result of the above-mentioned bug in the FTDNA software.

epp
02-16-2017, 09:22 PM
I’ve re-run my calculations from updated FTDNA data on the assumption that none of its basal R1a samples have Jewish ancestry. These now give the following ‘most likely’ estimates:
R1a1a1 point of origin close to the Netherlands - spreading North and East c. 3,500 BC
R1a1a1b point of origin close to the Czech Republic, followed by a fairly rapid expansion and development in R1a1a1b1a, centred on Poland, and spreading East into the Baltic states/Northern Russia and North into Sweden
No evidence of R1a1a1b (x R1a1a1b1) until it reaches a MRCA point of origin in Turkey (as R1a1a1b2) c. 2,300 BC, followed by a rapid expansion and development in all directions.
I have no particular opinion on these estimates, but they do prompt the following questions:
1. Might the estimated initial Eastward and Northward migrations of R1a1a1 have been influenced by R1b-L23, which my further analysis of FTDNA data estimates arrived from the South into the original R1a1a1 North West European homeland c. 4,000 BC?
2. Might R1a1a1b (x R1a1a1b1) have been part of a collapsing Yamna culture that was nearly decimated, and whose remnants later resurfaced in Turkey?, and if so -
3. Might any such collapse have been influenced by movement of (i) R1b-L151 Eastwards, and/or (ii) R1a1a1b1a Southwards, which the calculations both estimate to have occurred to some degree c. 2,500 BC?