PDA

View Full Version : New study on R1b Ht35 published by Lucotte



TigerMW
05-03-2013, 03:22 PM
I haven't read this yet but it appears to explain what ht35 is technically. Generally speaking, this is related to the R1b-L23xL51 folks found in the R1b M269+ U106- P312- project, also known as the "new ht35" project or now the "real R1b1a2*" project.

"Ht35 Y Chromosome Haplotype in Europe" by Lucotte
https://docs.google.com/file/d/0B1RIQRMwAdjpVVZRRUhDTE91M1U/edit?usp=sharing

I see that Peter Hrechdakian is listed as an author too. Here is the project background screen which has a little more explanation.
http://www.familytreedna.com/public/ht35new/default.aspx

Make sure to look at the frequency map. This looks like our best best bet to-date as far as a frequency chart for this group. As you know, I think diversity rather than frequency is of value for tracking an origin point, however, frequency definitely helps define the distribution and that can be compared with various cultural/folk movements.

It looks like the southern half of Italy has a lot of these people as well as the Balkan Peninsula, the Peloponnese and, of course, Anatolia. It looks like Romania, southern Poland and the Ukraine could be seen as an extension of the Balkans.

Unfortunately I don't think they covered over into Central Asia, not that there was a lot to find there anyway.

Rathna
05-03-2013, 03:52 PM
I haven't read this yet but it appears to explain what ht35 is technically. Generally speaking, this is related to the R1b-L23xL51 folks found in the R1b M269+ U106- P312- project, also known as the "new ht35" project or now the "real R1b1a2*" project.

"Ht35 Y Chromosome Haplotype in Europe" by Lucotte
https://docs.google.com/file/d/0B1RIQRMwAdjpVVZRRUhDTE91M1U/edit?usp=sharing

I see that Peter Hrechdakian is listed as an author too. Here is the project background screen which has a little more explanation.
http://www.familytreedna.com/public/ht35new/default.aspx

Make sure to look at the frequency map. This looks like our best best bet to-date as far as a frequency chart for this group. As you know, I think diversity rather than frequency is of value for tracking an origin point, however, frequency definitely helps define the distribution and that can be compared with various cultural/folk movements.

It looks like the southern half of Italy has a lot of these people as well as the Balkan Peninsula, the Peloponnese and, of course, Anatolia. It looks like Romania, southern Poland and the Ukraine could be seen as an extension of the Balkans.

Unfortunately I don't think they covered over into Central Asia, not that there was a lot to find there anyway.

Many thanks for the link. I saw that Bernard posted it on eng.molgen and said that Calabria (South Italy) gets the highest percentage of this haplogroup, thing I didn't doubt of. I'll read it soon.

Bernard
05-03-2013, 04:22 PM
These are the ht35 frequences in different european regions:
418
The maximum of frequences is in Calabria with a frequence of 23.1%.
This is the frequences map:
419

TigerMW
05-03-2013, 04:29 PM
3524 men were tested in Europe from 45 regions. 106 were R1b-ht35.

Thanks, Bernard. I'll delete the other thread so as not to confuse everyone.

alan
05-03-2013, 05:13 PM
I speed read this one. I notice that presumed ht35 is as high in northern Iran (but very low is southern Iran) as among Armenians. Unless these northern Iranians are really Armenians then I dont think we can link the high ht35 among north Iranians to a move from the Balkans. However northern Iran is pretty close to Armenia. Maybe they were mainly Armenians. Does anyone know? If the historical account of the move of Armenians from the Balkans is correct then I suppose that makes the whole peak around Armenia area dubious in terms of origin. I dont see an origin in the extreme south of Italy as very likely although clearly it had an impact there. Clearly L23 has had phases of moving about in the eastern and central Med. historically. It also confused by multiple empires that effected the east and central Med. Overall my eye gets drawn to the Balkans in this study. However, nothing is clear. Some variance calculations would have been useful

Rathna
05-03-2013, 05:19 PM
Some remarks about the paper:
1) This was called the Armenian haplogroup, and in fact each of us knew that near the Caucasus (also in Turkey) the R-L23 are above all Armenians, but Armenians are an Indo-European speaking people and their origin could be in the Balkans more than in the Caucasus.
2) This frequency map is in contrast with the other posted by Mikewww (from Busby) which had three places of greater diffusion (Alpine Region, the Caucasus, Central Asia) and certainly that South Italy has a so high percentage is noteworthy.
3) Lucotte says that the diffusion from Caucasus happened after the Last Glacial Maximum, thus he shouldn’t be used from the theorists who presupposed a recent diffusion of hg. R1b from East Europe or Middle East.
4) Where now R-L23 is pretty rare but there are its subclades (at percentages near to 100%) we shouldn’t search for the ancestor where we have so many grandsons.
5) This to say that the places where R-L23 is more diffused aren’t probably the places of origin. I always bet on a intermediate place, above all if it has all the pathway of the haplogroup, from R1b1 to R-P312* to R-U152 etc.

R.Rocca
05-03-2013, 06:12 PM
While it's not really about frequency, it is about splitting points. I've pointed out the small but dense amount of M269(xL23) in Calabria in the FTDNA Italy project which is interesting due to Myres finding strong numbers in the western Balkans (Montenegro and Serbia).

Rathna
05-03-2013, 07:01 PM
While it's not really about frequency, it is about splitting points. I've pointed out the small but dense amount of M269(xL23) in Calabria in the FTDNA Italy project which is interesting due to Myres finding strong numbers in the western Balkans (Montenegro and Serbia).

And they also have the rare YCAII=17-23, which isn’t limited to Calabria (see Ferrero from Piedmont), but recently I found also a Russian. I am following those people from many years and I know everything about them. But don’t forget that also Calabria is a mountainous region where ancient people probably survived, and, in spite of what is said about the Greek colonization, it is a place of ancient Italic languages settlements, demonstrated by toponyms. The Arbereshet are easily detectable, like Varipapa.

Jean M
05-03-2013, 08:53 PM
It looks like the southern half of Italy has a lot of these people as well as the Balkan Peninsula.

The southern half of Italy was settled by Greeks before the Roman empire sucked in the whole of the peninsula. It remained heavily Greek for long afterwards. I imagine that it was Italic earlier, but R1b-L23+ seems correlated with the "Balkan group" of IE languages: Greek, Armenian, Thracian etc.

Rathna
05-03-2013, 09:46 PM
The southern half of Italy was settled by Greeks before the Roman empire sucked in the whole of the peninsula. It remained heavily Greek for long afterwards. I imagine that it was Italic earlier, but R1b-L23+ seems correlated with the "Balkan group" of IE languages: Greek, Armenian, Thracian etc.

If this were true, we should have found in Southern Italy some R-L23 “ Balkan cluster”, but it isn’t so. The “Balkan cluster” found belong to Arbereshet (Ciulla) and Varipapa is R-Z2110+, not found so far in Italy and I think it won’t be found in the future. R-L23+ is at least 4% in Tuscany (1KGP) and much more in the Alpine zone, not colonized by Greeks. As I have said, the link with the Balkans and the Caucasus is much older, and it isn’t said from East to West rather than from West to East.

greystones22
05-03-2013, 09:49 PM
The southern half of Italy was settled by Greeks before the Roman empire sucked in the whole of the peninsula. It remained heavily Greek for long afterwards. I imagine that it was Italic earlier, but R1b-L23+ seems correlated with the "Balkan group" of IE languages: Greek, Armenian, Thracian etc.

Could the high frequency in the two southern Italy regions be attributed to some kind of social selection or biases? Were the Greek communities there doing something that could give their male offspring a better opportunity to have large families that stayed in the area?

There are some Greek speaking villages in southern Italy, but aren't they much more recent migrants?

greystones22
05-03-2013, 09:54 PM
If this were true, we should have found in Southern Italy some R-L23 “ Balkan cluster”, but it isn’t so. The “Balkan cluster” found belong to Arbereshet (Ciulla) and Varipapa is R-Z2110+, not found so far in Italy and I think it won’t be found in the future. R-L23+ is at least 4% in Tuscany (1KGP) and much more in the Alpine zone, not colonized by Greeks. As I have said, the link with the Balkans and the Caucasus is much older, and it isn’t said from East to West rather than from West to East.

Not necessarily, because Y chromosomes have a tendency to boom or bust in populations (aka drift).

AJL
05-03-2013, 11:00 PM
Armenians are an Indo-European speaking people and their origin could be in the Balkans more than in the Caucasus.

Sorry, but that's just preposterous, since we have M269* in the same area around the Caucasus, and since autosomal analysis, Y, and mtDNA all put Armenians solidly in and around Armenia, and since the L23 in this area are not just Armenians, they are also Assyrians, Jews, Kurds, northern Arabs, etc.

R.Rocca
05-04-2013, 12:45 AM
Not necessarily, because Y chromosomes have a tendency to boom or bust in populations (aka drift).

Guys, the Greeks only colonized the coasts in southern Italy and eastern Sicily. The Romans documented the names of the Italic tribes that still held the highlands at the time of their conquest. Besides, how is Calabrian L23 supposed to derived from Greece if it is more frequent in Calabria than in Greece? If anything, the Greeks probably reduced the amount of L23 and not the other way around.

alan
05-04-2013, 12:53 AM
Sorry, but that's just preposterous, since we have M269* in the same area around the Caucasus, and since autosomal analysis, Y, and mtDNA all put Armenians solidly in and around Armenia, and since the L23 in this area are not just Armenians, they are also Assyrians, Jews, Kurds, northern Arabs, etc.

It is an interesting question as to whether the Armenian L23 peak is due to the movement of that group from the Balkans that appears to be the ancient historians belief or whether it is far older than that. I will sit on the fence. L23XL51 and a smaller amount of M269* is a common pattern in both the Balkans and Armenians. I understand what you are saying about L23 being present among other groups in this area but at least some of them must have been later crossing of cultural barriers by male lineages. In addition the Armenians appear to have followed a similar path, albeit at a later time, as the Hittites who also are thought to have moved from the Balkans to Asia Minor and adjacent in the Bronze Age L23 may well have been associated with the latter too. From them and their large empire L23 could have spread widely and subsequenly been absorbed into many other ethnic groups after the dissapearance of the Hittite and other IE Anatolian peoples.

I am uncertain on this one but L23 does appear to have its largest block of elevated number is what could broadly be called the circumpontic region. Or at least the west, east and south sides of the Black Sea. As I have pointed out on other forums. the steppe and crimea part of Ukraine has seen almost total ethnic, linguistic and population change in the last 200 years or so, never mind the last 5000 years. So, we cannot really judge the steppe area of the Black Sea from modern populations. The question is whether this circumpontic distribution is ancient or whether the Balkan or Anatolia/Caucasus sections are late derivations from each other.

alan
05-04-2013, 01:08 AM
I would still feel that the main elevated block is the Balkans and the south/south-east side of the Black Sea. I just cannot see the south Italian peak as anything other than an offshoot when the overall phylogeny of early R1b and its distibrution is considered. It would be different if the south of Italy had a really convincing range of R1b upstream of L23. That would have been an unexpected jackpot and would maybe have identified the R1b refuge in an unexpected place.

Anyway, the date of L23XL51 has to be considered to understand its spread. What is the current idea of the date. Is it something like 6000 years old overall??

alan
05-04-2013, 01:31 AM
The Albabian peak is interesting too in that it had already been noted for its high L23 and M269 counts, the only place in the Balkans where M269* and L23* are high both proportionately and in absolute terms. This is the likely link to Italy as Albania lies directly opposite southern Italy. Many consider the Albanians to by the remnants of the Illyrians.

AJL
05-04-2013, 03:42 AM
L23 does appear to have its largest block of elevated number is what could broadly be called the circumpontic region.

Okay, that, I'll buy. To put the origin squarely and conclusively in the Balkans, though, seems not to fit the phylogeny of Q/R1a/R2 nor R1b1(a/2)*, though with some R-M198*/R-M417* in Eastern Europe, it is still possible.

{EDIT}
I see Silesian makes the same point in this thread too:

http://www.anthrogenica.com/showthread.php?820-R1b-Early-Branching-Phylogeny-(SNP-based-family-tree)&p=5301&viewfull=1#post5301

Sorry for the cross-thread headache.

Jean M
05-04-2013, 10:44 AM
To put the origin squarely and conclusively in the Balkans, though...

Sorry if I gave the impression that I thought L23 originated in the Balkans. I have no idea where it originated. Or at least I cannot decide between various options.

L23 is found in eastern Europe and the Caucasus, Turkey and Circum-Urals (Myres 2010), but this is likely to be almost all all Z2105. Until we can get a true map of L23* and most importantly L23* in aDNA, we are groping in the dark. My guess would be that L23* originated somewhere en route between the northern Zagros and the Sea of Marmara, but who knows? The L23 in Iran appears to have arrived there with Armenians and Assyrians, so I doubt if that is the origin point.

I observed that we find L23 (probably Z2105) in present-day populations speaking Greek and Armenian. These two languages are part of what linguists call the Balkan group, since they appear to have developed into proto form in the Balkans before the ancestor of Phrygian and Armenian split away to enter Anatolia. Armenian did not enter what is now the Armenian plain until the 6th century BC. Both Greek and Armenian speakers appear to have absorbed a lot of locals in the regions where they settled, but L23 seems a common thread between them. However we do not know if L23 was already in Thrace (from early dairy farmers) when the ancestors of Greeks and Armenians arrived in it, or whether the said ancestors arrived with it from Cucuteni/Yamnaya.

The L23 in Assyrians needs to be explained. The most economical explanation might lie in some L23 input into the Halaf culture (along with V88?) But again we are just groping in the dark.

alan
05-04-2013, 11:06 AM
Okay, that, I'll buy. To put the origin squarely and conclusively in the Balkans, though, seems not to fit the phylogeny of Q/R1a/R2 nor R1b1(a/2)*, though with some R-M198*/R-M417* in Eastern Europe, it is still possible.

{EDIT}
I see Silesian makes the same point in this thread too:

http://www.anthrogenica.com/showthread.php?820-R1b-Early-Branching-Phylogeny-(SNP-based-family-tree)&p=5301&viewfull=1#post5301

Sorry for the cross-thread headache.

I totally agree. Nothing is clear once we are above M269, M73 and V88. There is just a lack of a convincing concentration of ancestral forms above those branches. Its just very diffuse. There is also the problem of incredible population change since those branches. SE Europe, the steppes, Anatolia etc have not exactly had a quiet time historically. I favour some sort of ancient position that was not in the main hop off points for the earliest farmers to Europe and perhaps even contained within a group that experienced farming rather later than the first flourish of farming. That seems to be the story etched on the genes with the age of M269 (well L23 really) and V88 being a couple of millenium later than the spread of farming into Europe and much younger than farming in its area of origin. It seems to have been hiding when other clades like G were colonising Europe. So, its probably safe to say it wasnt siting on the Levant coast in the early Neolithic. There were substrate groups of hunter-fishers of a very similar nature all around the Black Sea shore from north-west Anatolia to the Crimea and perhaps other areas of the coast (the coast has been lost in many areas). They were apparently incorporated into the pstoralist groups. Its also the area where advance cattle pastoralism arose. I have a couple of books on this but I found this fairly recent paper online that is interesting

http://www.academia.edu/1581680/The_Emergence_of_Neolithic_Life_in_South_and_East_ Marmara_Region


I

alan
05-04-2013, 11:18 AM
I was just thinking L23* (I just use that as shorthand) seems to have that classic peak at the edge of its main distribution. Another thing I would say about the L23* map in the new report is its broad brush. For example we know from the recent Bulgaria study that it is MUCH higher at the Black Sea coast than inland or compared to Bulgaria as a whole. We also know from another recent paper it is also elevated in Moldovia, a near-Black Sea country linked to Romania culturally and linguistically. The broadbrush sampling of this paper does blur this Black Sea concentration somewhat. I would still be most comfortable with the term circumpontic for this group even if it is a bit of a cop out. Clearly today it is more elevated along the north, south and south-east coasts of the Black Sea as well as the peaks in Albania and Italy that this paper does identify. My impression from this and other papers is that it is not so high in the mountanous inland areas of the Balkans than around its coastal fringe.

newtoboard
05-04-2013, 11:42 AM
I speed read this one. I notice that presumed ht35 is as high in northern Iran (but very low is southern Iran) as among Armenians. Unless these northern Iranians are really Armenians then I dont think we can link the high ht35 among north Iranians to a move from the Balkans. However northern Iran is pretty close to Armenia. Maybe they were mainly Armenians. Does anyone know? If the historical account of the move of Armenians from the Balkans is correct then I suppose that makes the whole peak around Armenia area dubious in terms of origin. I dont see an origin in the extreme south of Italy as very likely although clearly it had an impact there. Clearly L23 has had phases of moving about in the eastern and central Med. historically. It also confused by multiple empires that effected the east and central Med. Overall my eye gets drawn to the Balkans in this study. However, nothing is clear. Some variance calculations would have been useful

Why would you assume that they are Armenians? most studies that have tested the Armenians/Assyrians of Iraq have stated they tested them and didn't try to pass them off as ethnic Iranians? I recall ht35 being just as common in SW Iran(Luristan) and the Caspian coast as in NW Iran(the area closest to Armenia) anyways.

AJL
05-04-2013, 12:16 PM
Sorry if I gave the impression that I thought L23 originated in the Balkans.

Jean: No need to apologize. It was Rathna who, at least to me, seemed to be proposing this.

AJL
05-04-2013, 12:21 PM
http://www.academia.edu/1581680/The_Emergence_of_Neolithic_Life_in_South_and_East_ Marmara_Region


Interesting indeed, thanks. One wonders if flooding in this area may have sent either waves (if you will pardon the pun), or a single wave, of people both east and west at the right time to have contributed significantly to current R1b.

AJL
05-04-2013, 12:23 PM
Why would you assume that they are Armenians?

He is not assuming this. The conjunction "as" is implying "Northern Iran" and "Armenians" are parallel, not identical.

alan
05-04-2013, 12:30 PM
Why would you assume that they are Armenians? most studies that have tested the Armenians/Assyrians of Iraq have stated they tested them and didn't try to pass them off as ethnic Iranians? I recall ht35 being just as common in SW Iran(Luristan) and the Caspian coast as in NW Iran(the area closest to Armenia) anyways.

This paper did show a vastly higher total in northern than southern Iran (I have no idea what the ethnicities were - I would hope that they did check ethnic minorities in their sampling). I only raised Armenians and Hittites because many think they link the Balkans and Anatolia/Armenia. I actually have never been 100% convinced by the historic Balkans-Anatolia moves and I only raised them because many do believe in this sort of move. There are also considerable links between northern Anatolia and the Balkans in the Mesolithic and Neolithic. Basically I was looking for a common denomenator that links both areas and the Mesopotamian and Assyrian link is harder to see as a common denomenator between the SE European, Anatolian and SW Asian areas. However, noone knows really. There was a Mesopotanian link with Maykop at the origin of the Circumpontic Metallurgical Provence c. 4000-3000BC so I wouldnt rule out that sort of link either although its more hazy to define.

This study is broad brush and recent studies of Bulgaria and Moldovia do suggest that national averages such as in the recent report mask the Black Sea concentration of L23*. The further concetration in Albania and south Italy is not entirely new although much more striking than previously thought. This again suggest a coastal distribution - the Bulgaria paper already having demostrated that L23* massively fades inland from the Black Sea. Armenia as well in of course close to the Black Sea.

The term circumpontic should not be seen as too literal though - it takes in lots of land some distance away from the Black Sea including northern Mesopotamia and the steppes. Whatever its origin it does look to pick up its frequency in the circumpontic zone as the Black Sea is approached and it seems to have found some success in that zone whether or not it was the origin. So much of the Neolithic shore has been lost that we may be missing something important. A recent paper suggests that there are hints that the northern shoreline of the Black Sea also once featured Neolithic groups with links to the south but this is largely submerged (the only likely area for fruitful land based future research being the Crimean where the ancient shore has changed little). its still all a mystery and you are right in the sense that ethnic labels are not helpful and pretty speculative anyway given the passing of 6000 years or so since L23*.

alan
05-04-2013, 01:19 PM
Another obvious observation is that in SE Europe L23* clearly is higher in the areas where Slavic intursions were less and falls away very roughy where Slavic languages are spoken in the western Balkans north of Albania. Its not a perfect correlation but its pretty strong (Moldovia I believe speak Romanian has a lot of L23* too). So, one thing that must be borne in mind is that the modern European distribution has likely been significantly altered by the Slavic expansion and may have been somewhat different in pre-Slavic expansion times.

alan
05-05-2013, 10:43 PM
I must say though that the conclusions and dating in this report look ropey. I dont really understand the reasoning to pinpoint Armenia as the origin point. That seems to be purely based on that area being the highest non-European frequency. Is there anything to be learned from the haplotypes in this paper? Variance by area? Its also a pity that an overall map of projected L23* in Asia wasnt included.

newtoboard
05-06-2013, 01:30 AM
I must say though that the conclusions and dating in this report look ropey. I dont really understand the reasoning to pinpoint Armenia as the origin point. That seems to be purely based on that area being the highest non-European frequency. Is there anything to be learned from the haplotypes in this paper? Variance by area? Its also a pity that an overall map of projected L23* in Asia wasnt included.

There isn't any. Maybe because all the people calling the Armenian highlands the cradle of civilization and calling it the IE homeland(despite claims against both) has started to have an impact on research.

Silesian
05-06-2013, 03:20 AM
I must say though that the conclusions and dating in this report look ropey. I dont really understand the reasoning to pinpoint Armenia as the origin point. That seems to be purely based on that area being the highest non-European frequency. Is there anything to be learned from the haplotypes in this paper? Variance by area? Its also a pity that an overall map of projected L23* in Asia wasnt included.

It's natural to want to know where are R1b ancestors are from and the paths they took. Even how we are related in the grand scheme of things to our branched neighbors like R2 and Q[for me anyway]. Armenian is technically a Indo-European language and all R1b who post here on the forum share ancestral snp's like R-M269 and L23 with them. It would be nice to see some Armenians with ydna R1b actually post here and contribute. However, I agree it looks a little short on some specifics like variance by area.

ADW_1981
05-06-2013, 03:32 AM
Guys, the Greeks only colonized the coasts in southern Italy and eastern Sicily. The Romans documented the names of the Italic tribes that still held the highlands at the time of their conquest. Besides, how is Calabrian L23 supposed to derived from Greece if it is more frequent in Calabria than in Greece? If anything, the Greeks probably reduced the amount of L23 and not the other way around.

Not necessarily. E-V13 likely represents the only European branch of the oldest Levantine populations. On the flipside southern Greece is a very rich in R1b population. If anything, population movements from the north east and east into Greece have lowered the R1b frequency in Greece and the surrounding area. Just my opinion though.

http://www.biomedcentral.com/1471-2148/11/69

Something to consider is that this paper only focuses on DYS393 = 12. It does not look at SNPs at all, and the Albanian/Greek cluster won't even be recognized as a ht35 haplotype even if it is one. The Armenian refugia is a rubbish theory. There is enough evidence here that may indicate an ancient spread of R1b into West Asia from the Balkans.

TigerMW
05-06-2013, 04:15 AM
Guys, the Greeks only colonized the coasts in southern Italy and eastern Sicily. The Romans documented the names of the Italic tribes that still held the highlands at the time of their conquest. Besides, how is Calabrian L23 supposed to derived from Greece if it is more frequent in Calabria than in Greece? If anything, the Greeks probably reduced the amount of L23 and not the other way around.

I don't know if L23xL51 moved from Greece to Italy or Italy to Greece [EDIT to correct slip of the pen] or something else. I'm just trying to follow along.

However, in terms of logic, we can not say that because something is more frequent in Calabria than Greece that the movement was that direction. Of course, we can't say the converse as well. This is where various diversity measures would be helpful if we had enough deep data. I'm just saying I would consider frequency is a non-issue in the direction of migration issue. We know there is a presence in both regions.

TigerMW
05-06-2013, 04:32 AM
...
Something to consider is that this paper only focuses on DYS393 = 12. It does not look at SNPs at all, and the Albanian/Greek cluster won't even be recognized as a ht35 haplotype even if it is one.

Agreed, this is problematic. I wish we could get the R1b ht35 project admin to post here on this. The 393=13 L23* folks were more prominent in the Assyrian and Alawite folks versus the Armenian L23*. You can make some interesting theories about that but some of that could be defeated or diminished by thorough Z2103 and Z2105 testing.


The Armenian refugia is a rubbish theory. There is enough evidence here that may indicate an ancient spread of R1b into West Asia from the Balkans.

What is the Armenian refugia theory and why it is rubbish?

Rathna
05-06-2013, 07:14 AM
I don't know if L23xL51 moved from Greece to Italy or Greece to Italy or something else.

Sigmund Freud: lapsus calami

Rathna
05-06-2013, 07:21 AM
The 393=13 L23* folks were more prominent in the Assyrian and Alawite folks versus the Armenian L23*.


The Balkan R-L23 (Z2105+) with DYS393=13 is well identified like the "Balkan cluster" (not only this value but also DYS459a=8, DYS446=14 or more etc.), and I have said elsewhere that this haplotype lacks in Italy except for those who are of Arbereshe origin and it excludes that Italian R-L23-s derive from the Balkans. In the past I carried "proofs" of the other way around. It'd be interesting to test for SNPs Italian Risso, etc.
Also Varipapa with his R-L23/Z2110+ is of Arbereshe origin and this SNP hasn't been found so far in Italy.

alan
05-06-2013, 11:34 AM
Is there actually and Armenia refugia theory in genetic terms? I know it is one of the minority propositions for IE languages but I cannot recall it being discussed in terms of refugia. I understand the idea of an ice age refuge around the south Caspian but that is not Armenia per se. That of course brings us to the issue of the dating suggested in this paper. I take it this is all based on the out of date ideas using the Zhiv rates etc. I do understand from past work it was shown that L23XL51 is a lot lower in variance in Europe compared to outside Europe with the exception of the western Black Sea coast. However, I also recall this being challenged.

I think whatever the story ir needs to have some plausible link to archaeological evidence and of course establishing age of the clade by region is important in this too or there is no solid ground to use. The evidence of flow from east to west tends to be much stronger in the early archaeological record than the reverse although there are exceptions and it gets a whole lot more complex in later prehistory and historic times with all sorts of empires. I wonder if anyone with map graphic skills will produce a map based on this and any other data that also includes more of SW Asia, central Asia etc and maybe refines it with data such as the Bulgaria and Moldova studies. The map in this study makes an excellent start but could be expanded and refined. Eupedia seem to like making maps so maybe one will appear over there.

One other observation that I would like to float is that the majority of groups with raised L23* seem to be non-Slavic IE speakers (or where a non-IE language was a much later elite imposed language over an IE population like Bulgaria and Hungary). Of course then there are the Armenians and Iranians. I am not playing down the non-IE aspect or trying to say it is impossible that L23* was originally IE but it is largely today in areas that are or formerly were IE speaking. You could probably say the same about the various groups in central Asia where Turkic is a superstate originating from beyond the R1b world.

alan
05-06-2013, 11:43 AM
Here is a question for people more knowledgeble on DNA than me - is L23* the only common denomenator of a significant size between all these SE European and Asian peoples that now speak IE (Greeks, Armenians, Albanians, Romanians, Moldovians. I would ignore the more purely or partly Slavic speaking groups because they have a compicated history that involved multiple IE layers and R1a is clearly their most likely IE marker. Is there any other clades shared by them all at above trace levels. Because if there is no other significant shared clade above trace levels then EITHER L23* must be the common IE thread OR the idea that IE spread should be echoed by a yDNA trail is wrong.

This Wiki page probably makes a starting point for discussing this

http://en.wikipedia.org/wiki/Y-DNA_haplogroups_by_ethnic_group

alan
05-06-2013, 12:00 PM
The Ararat study noted 'only low frequencies of haplogroup R1a1*-M198, which has been associated with the Indo-Aryan expansions,36 were observed in Ararat Valley (0.9%), Gardman (5.2%) and Sasun (0.9%)'. While not absent that would seem to indicate areas of people speaking an IE language with very small amounts of R1a.I also note that Calabrians and south Italians in general have almost no R1a.

However a serious problem is that the Slavs made some impact even in many non-Slavic speaking areas and even effected non-Slavic languages like Romanian etc. Other groups have had a complex history - Albabians and Serbs etc. Northern Greece too had a major Slavic element. So its very very hard to find groups in the L23* area that are entirely free of some Slavic contribution above the trace level. It is interesting though that the very high L23XL51 areas in south Italy and Armenia at the fringes of the main L23XL51 block had areas with virtually no R1a. History elsewhere just seems too messy to draw absolute conclusions. Nonetheless it does show that its not just in the west of Europe that IE speakers with very little R1a existed. IF the Armenian-Balkan story is correct (I have no opinion on this) and the Armenians yDNA preserves the pattern of later prehistoric Balkans peoples then it would suggest that R1a was very uncommon in the Balkans before the Slavic expansion.

alan
05-06-2013, 12:59 PM
This Wiki page gives a summary of the long history of Albanians migrating to southern Italy.

http://en.wikipedia.org/wiki/Arb%C3%ABresh%C3%AB_people

I am not saying it explains the strong Albania-south Italy L23 link entirely but it would by silly not to factor this in.

Reharding the Albanians of Albania it is interesting that a number of linguists say the Albanian language indicates the Albanians were once non-coastal and perhaps somewhat more northern and eastern in location than today. Of course this is the subject of much dispute.

newtoboard
05-07-2013, 11:39 AM
Is there actually and Armenia refugia theory in genetic terms? I know it is one of the minority propositions for IE languages but I cannot recall it being discussed in terms of refugia. I understand the idea of an ice age refuge around the south Caspian but that is not Armenia per se. That of course brings us to the issue of the dating suggested in this paper. I take it this is all based on the out of date ideas using the Zhiv rates etc. I do understand from past work it was shown that L23XL51 is a lot lower in variance in Europe compared to outside Europe with the exception of the western Black Sea coast. However, I also recall this being challenged.

I think whatever the story ir needs to have some plausible link to archaeological evidence and of course establishing age of the clade by region is important in this too or there is no solid ground to use. The evidence of flow from east to west tends to be much stronger in the early archaeological record than the reverse although there are exceptions and it gets a whole lot more complex in later prehistory and historic times with all sorts of empires. I wonder if anyone with map graphic skills will produce a map based on this and any other data that also includes more of SW Asia, central Asia etc and maybe refines it with data such as the Bulgaria and Moldova studies. The map in this study makes an excellent start but could be expanded and refined. Eupedia seem to like making maps so maybe one will appear over there.

One other observation that I would like to float is that the majority of groups with raised L23* seem to be non-Slavic IE speakers (or where a non-IE language was a much later elite imposed language over an IE population like Bulgaria and Hungary). Of course then there are the Armenians and Iranians. I am not playing down the non-IE aspect or trying to say it is impossible that L23* was originally IE but it is largely today in areas that are or formerly were IE speaking. You could probably say the same about the various groups in central Asia where Turkic is a superstate originating from beyond the R1b world.

Maybe a study is needed on populations of the SW Caspian. I think the Talysh possess R1a*, R1a, R1b and R2. Do you know anything about their R1b types? I agree with the Armenian theory being bad.

razyn
05-07-2013, 01:02 PM
This Wiki page probably makes a starting point for discussing this

http://en.wikipedia.org/wiki/Y-DNA_haplogroups_by_ethnic_group

That page has some very strange stats. Some of the rows add up to well over 100% (108% of the Basque sample, 144% of Ossetians), and some only around 10%, 40%, whatever interested the researcher I guess. 78% of the French, 65% of the English. A lot of papers are cited, and many are probably useful ones on their specific topics, but as a table it's pretty bogus.

Silesian
05-09-2013, 06:00 AM
Here is a question for people more knowledgeble on DNA than me - is L23* the only common denomenator of a significant size between all these SE European and Asian peoples that now speak IE (Greeks, Armenians, Albanians, Romanians, Moldovians. I would ignore the more purely or partly Slavic speaking groups because they have a compicated history that involved multiple IE layers and R1a is clearly their most likely IE marker. Is there any other clades shared by them all at above trace levels. Because if there is no other significant shared clade above trace levels then EITHER L23* must be the common IE thread OR the idea that IE spread should be echoed by a yDNA trail is wrong.

This Wiki page probably makes a starting point for discussing this

http://en.wikipedia.org/wiki/Y-DNA_haplogroups_by_ethnic_group

Of interest the Ossetian samples are probably L23x51{2000 Rosser 42.6%- N47} and perhaps similar to Georgian/X-Kakhetians
http://www.familytreedna.com/public/Georgia/default.aspx?section=yresults
However some look very different, compared to Georgian_ National Clans project_India subcontinent DNA Project, and Hui sample.
223828 Kazak R1b1a2a1 12 21 14 11 11-15 12 12 12 13 14 28 17 9-10 11 11 25 15 19 28
259018 Kakheti-- R1b1a2 11 24 14 11 12-14 12 12 12 13 13 29 19 9-10 11 11 24 15 19 27
204505 Pakistan. R1b1a2 12 24 14 10 11-14 12 13 12 13 13 29 15 9-10 11 11 26 15 19 31
Chinese Hui.----- R1b1a2 12 24 15 10 11-14 12 12 12 13 13 29
http://www.familytreedna.com/public/Bashqort_Clans/default.aspx?section=yresults
It would be interesting to see some samples from Talysh/Gilaki/Lur North-Western Iran to see if they resemble any of the L23X51 from Northern Pakistan/Gujarit/Bengal.
http://www.familytreedna.com/public/India/default.aspx?section=yresults

Rathna
05-09-2013, 09:40 AM
Chinese Hui.----- R1b1a2 12 24 15 10 11-14 12 12 12 13 13 29


I have spoken many times of this Hui who matches me (Gioiello Tognoni) and a far relative of mine (Giancarlo Tognoni) not only in these 12 markers but also in the others (take present that YHRD H4 must be diminished of one).
My theory was that he descended from a Roman Soldier of the Crassus’ Army defeated at Charrae on 53 BC and after deported at Merv and who fought with White Huns against Chinese (all documented by Chinese historians).
The Han Richard Rocca spoke about from Beijing I suppose he descends from the same Roman soldier. He is the unique on 1KGP negative for Z2106 to Z2110. Of course it would be a proof to test me for these SNPs.

4 15 13 29 24 10 13 12 11,14 -1 -1 -1 -1 -1 -1 -1 -1 -1 -1 -1 -1 -1 -1 >>
3 15 13 29 24 10 13 12 11,14 12 11 15 19 14 17 23 11 -1 -1 -1 -1 -1 -1 >>
3 15 13 29 24 10 13 12 11,14 12 12 -1 -1 -1 -1 -1 -1 -1 -1 -1 -1 -1 -1 >>
2 15 13 29 24 10 13 12 11,14 12 13 15 -1 -1 -1 -1 -1 -1 -1 -1 -1 -1 -1 >>
1 15 13 29 24 10 13 12 11,14 12 12 15 19 15 17 23 12 -1 -1 -1 -1 -1 -1 >>
1 15 13 29 24 10 13 12 11,14 12 12 15 19 15 19 23 12 -1 -1 -1 -1 -1 -1 >>
1 15 13 29 24 10 13 12 11,14 12 13 -1 -1 -1 -1 -1 -1 -1 -1 -1 -1 -1 -1 >>
1 15 13 29 24 10 13 12 11,14 12 11 14 19 17 17 24 12 -1 -1 -1 -1 -1 -1 >>
1 15 13 29 24 10 13 12 11,14 12 13 15 19 15 16 23 12 -1 -1 -1 -1 -1 -1 >>
1 15 13 29 24 10 13 12 11,14 12 13 15 19 15 16 23 12 -1 -1 -1 -1 -1 -1 >>
1 15 13 29 24 10 13 12 11,14 12 13 15 19 16 16 24 12 -1 -1 -1 -1 -1 -1 >>
1 15 13 29 24 10 13 12 11,14 12 13 15 19 16 17 23 11 -1 -1 -1 -1 -1 -1 >>
1 15 13 29 24 10 13 12 11,14 13 12 15 -1 -1 -1 -1 -1 -1 -1 -1 -1 -1 -1 >>

2 of 645 Prague, Czech Republic [Czech] Eurasian - European - Eastern European Europe
2 of 255 Rio Grande do Sul, Brazil [European] Eurasian - European - Western European Latin America
1 of 587 Buenos Aires, Argentina [Admixed] Admixed Latin America
1 of 300 Brasilia, Brazil [Admixed Brazilian] Admixed Latin America
1 of 637 Rio de Janeiro, Brazil [Admixed] Admixed Latin America
1 of 270 East Tyrol, Austria [Tyrolean] Eurasian - European - Western European Europe
1 of 1079 Australia [European] Eurasian - European - Western European Oceania / Australia
1 of 205 Bahia, Brazil [Admixed Brazilian] Admixed Latin America
1 of 220 East Croatia, Croatia [Croatian] Eurasian - European - South-Eastern European Europe
1 of 97 Reggio di Calabria, Italy [Italian] Eurasian - European - South-Eastern European Europe
1 of 101 Athens, Greece [Greek] Eurasian - European - South-Eastern European Europe
1 of 130 Modena, Italy [Italian] Eurasian - European - Western European Europe
1 of 91 Caceres, Spain [Spanish] Eurasian - European - Western European Europe
1 of 99 Strasbourg, France [French] Eurasian - European - Western European Europe
1 of 245 Rio de Janeiro, Brazil [European] Eurasian - European Latin America
1 of 281 Munich, Germany [German] Eurasian - European - Western European Europe
1 of 118 Ningxia, China [Hui] East Asian - Sino-Tibetan - Chinese Asia
1 of 942 Gdansk, Poland [Polish] Eurasian - European - Eastern European Europe
1 of 427 Córdoba, Argentina [European] Eurasian - European Latin America

Rathna
05-09-2013, 10:03 AM
1 15 13 29 24 10 13 12 11,14 12 13 15 19 15 16 23 12 -1 -1 -1 -1 -1 -1 >>
1 15 13 29 24 10 13 12 11,14 12 13 15 19 15 16 23 12 -1 -1 -1 -1 -1 -1 >>

1 of 637 Rio de Janeiro, Brazil [Admixed] Admixed Latin America
1 of 130 Modena, Italy [Italian] Eurasian - European - Western European Europe

3 15 13 29 24 10 13 12 11,14 12 12 -1 -1 -1 -1 -1 -1 -1 -1 -1 -1 -1 -1 >>
1 15 13 29 24 10 13 12 11,14 12 12 15 19 15 17 23 12 -1 -1 -1 -1 -1 -1 >>
1 15 13 29 24 10 13 12 11,14 12 12 15 19 15 19 23 12 -1 -1 -1 -1 -1 -1 >>

1 of 300 Brasilia, Brazil [Admixed Brazilian] Admixed Latin America
1 of 220 East Croatia, Croatia [Croatian] Eurasian - European - South-Eastern European Europe
1 of 118 Ningxia, China [Hui] East Asian - Sino-Tibetan - Chinese Asia
1 of 514 Gdansk, Poland [Polish] Eurasian - European - Eastern European Europe
1 of 91 Caceres, Spain [Spanish] Eurasian - European - Western European Europe

1 15 13 29 24 10 13 12 11,14 12 12 15 19 15 17 23 12 -1 -1 -1 -1 -1 -1 >>

1 of 300 Brasilia, Brazil [Admixed Brazilian] Admixed Latin America

1 15 13 29 24 10 13 12 11,14 12 12 15 19 15 19 23 12 -1 -1 -1 -1 -1 -1 >>

1 of 220 East Croatia, Croatia [Croatian] Eurasian - European - South-Eastern European Europe

Then the Hui is amongst the three signed
3 15 13 29 24 10 13 12 11,14 12 12 -1 -1 -1 -1 -1 -1 -1 -1 -1 -1 -1 -1 >>

Silesian
05-20-2013, 03:37 PM
I have spoken many times of this Hui who matches me (Gioiello Tognoni) and a far relative of mine (Giancarlo Tognoni)......

Interesting place Italy.

Here is one who matches me at 12, from Schleswig-Holstein. Exact at 12 but extended we differ, mostly on the fast markers, which is why he belongs to 7.0 [L21+] P312+ M222- Extended and chosen from 37.
8/8 slow marker match[mutation rate]

393-12-0.00076%
426-12-0.00008%
388-12-0.00022%
392-13-0.00052%
455-11-0.00016%
454-11-0.00016%
437-15-0.00099%
438-12-0.00055%

13/17 medium match [mutation rate]

390-24-0.0031%
19-14-0.00151
391-10-0.00226%
385-11-14-0.00226%
439-11-0.00477%[/I]
389I-14-0.00186%
389II-31-0.00242%
459-9-10-0.00132%
447-25-0.00264%
448-19-0.00135%
YCAII-19-23-0.00151%-*DYS394
3 markers 1 step.
460-11-10
Gata-H4-11-10
442-11-12

2/10 fast match
464-b-c-15-16-0.00566%

Rathna
05-20-2013, 07:59 PM
Interesting place Italy.

Here is one who matches me at 12, from Schleswig-Holstein. Exact at 12 but extended we differ, mostly on the fast markers, which is why he belongs to 7.0 [L21+] P312+ M222- Extended and chosen from 37.
8/8 slow marker match[mutation rate]

393-12-0.00076%
426-12-0.00008%
388-12-0.00022%
392-13-0.00052%
455-11-0.00016%
454-11-0.00016%
437-15-0.00099%
438-12-0.00055%

13/17 medium match [mutation rate]

390-24-0.0031%
19-14-0.00151
391-10-0.00226%
385-11-14-0.00226%
439-11-0.00477%[/I]
389I-14-0.00186%
389II-31-0.00242%
459-9-10-0.00132%
447-25-0.00264%
448-19-0.00135%
YCAII-19-23-0.00151%-*DYS394
3 markers 1 step.
460-11-10
Gata-H4-11-10
442-11-12

2/10 fast match
464-b-c-15-16-0.00566%

Of course your sample doesn’t fit at all. You have chosen, amongst the thousands of people tested, that one who matches closely you but belongs to a different haplogroup. The Hui who matches me is one of the few who matches me all over the world, the others come all from Europe or from South America where Italians (for their Y) are even more than the Italians of Italy. No close match to this Hui has appeared so far in Asia, demonstrating, I think, that this haplotype is clearly of European origin. My markers value, unlike yours, aren’t close to the modal but very rare taken together(DYS19=15, DYS391=10, DYS456=15, DYS458=16 etc.).
If it is true that the other Chinese Han who is R-Z2105 (and probably genetically close to this Hui) is negative for Z2106-Z2110 (the unique found so far for what Richard Rocca says), I have said that only a test which demonstrates that me too are the same could link me for sure to these Chinese. But the 1KGP has already demonstrated that at that level of SNPs were found Tuscans and one Hui.

Hope you understood that my analyses could be wrong, but are made by knowledge and aren’t invalidated by malice and bad conscience.

Silesian
05-20-2013, 08:07 PM
Of course your sample doesn’t fit at all...Hope you understood that my analyses could be wrong, but are made by knowledge and aren’t invalidated by malice and bad conscience.

No problem thanks for having a look. :)

R.Rocca
05-20-2013, 08:28 PM
Of course your sample doesn’t fit at all. You have chosen, amongst the thousands of people tested, that one who matches closely you but belongs to a different haplogroup. The Hui who matches me is one of the few who matches me all over the world, the others come all from Europe or from South America where Italians (for their Y) are even more than the Italians of Italy. No close match to this Hui has appeared so far in Asia, demonstrating, I think, that this haplotype is clearly of European origin. My markers value, unlike yours, aren’t close to the modal but very rare taken together(DYS19=15, DYS391=10, DYS456=15, DYS458=16 etc.).
If it is true that the other Chinese Han who is R-Z2105 (and probably genetically close to this Hui) is negative for Z2106-Z2110 (the unique found so far for what Richard Rocca says), I have said that only a test which demonstrates that me too are the same could link me for sure to these Chinese. But the 1KGP has already demonstrated that at that level of SNPs were found Tuscans and one Hui.

Hope you understood that my analyses could be wrong, but are made by knowledge and aren’t invalidated by malice and bad conscience.

Thanks for the reminder. I just took a look at his sequence and he is indeed L277-. It looks like he was mislabeled at some point along the way as L277+.

Silesian
05-20-2013, 08:48 PM
IF the Armenian-Balkan story is correct (I have no opinion on this) and the Armenians yDNA preserves the pattern of later prehistoric Balkans peoples then it would suggest that R1a was very uncommon in the Balkans before the Slavic expansion.

No where else in Asia is there such a sudden drop in L23x51, especially the region around the Armenia/Assyria/ and Northwest Iran.

Meanwhile in Crete we can compare Lasithi Plateau with Armenian plateau, even though though the sample set is small here are some conclusions from dienekes, bolg.


The most outstanding differences were observed in haplogroups J2 and R1, with the predominance of haplogroup R lineages in the Lasithi Plateau and of haplogroup J lineages in the more accessible regions of the island. Y-STR-based analyses demonstrated the close affinity that R1a1 chromosomes from the Lasithi Plateau shared with those from the Balkans, but not with those from lowland eastern Crete. In contrast, Cretan R1b microsatellite-defined haplotypes displayed more resemblance to those from Northeast Italy than to those from Turkey and the Balkans.
http://dienekes.blogspot.ca/2007/01/y-chromosomes-from-lasithi-crete.html


Crete (Lasithi Plateau) (N=41)
R1b -- 36.6% (R1b1a2 = 29%)
R1a -- 19.5%
J2 -- 9.8%
G2a -- 7.3%
I -- 7.3%
T -- 7.3%
E1b -- 4.9%
Q -- 4.9%
J1 -- 2.4%

Silesian
05-21-2013, 12:36 AM
Of course your sample doesn’t fit at all....

This is very strange, how can Nazarov be a genetic distance of 27 out of 37 and still be in the same L23+? This is a greater distance than my exact Shleswig-Holstein [L21+] P312+ M222-] match at 12, which is only distant by 19 out of 37?

AJL
05-21-2013, 01:25 AM
This is very strange, how can Nazarov be a genetic distance of 27 out of 37 and still be in the same L23+?

Precisely because it is very old.

Rathna
05-21-2013, 02:49 AM
Precisely because it is very old.

But I’d add my three principles, no one took in consideration. Mutations happen 1) around the modal 2) there is a convergence to the modal as time passes 3) sometime a mutation goes for the tangent. The combination of these principles makes that Nazarov has a GD of 27 out 37 as to you and others much less, but only because Nazarov belongs to a different line of R-L23 and others to your same line.
I wrote about an Indian R1a1a, then belonging to a recent subclade as to Western European R1a-M420, but with so many mutations as to the same Western European subclades and calculated its age with the current method obtaining a time similar to the infamous Zhivotovsky rate.

For this the calculation of the other infamous Anatole Klyosov based upon the 22 slowest markers merits attention and I have used it in my assessment of the R1 subclades. These slow mutating markers suffered less than the fastest ones these three principals, the mutations are above all of 1 plus or minus, and this demonstrates that there haven’t been mutations around the modal.

TigerMW
06-03-2013, 06:37 PM
But I’d add my three principles, no one took in consideration. Mutations happen 1) around the modal 2) there is a convergence to the modal as time passes 3) sometime a mutation goes for the tangent. The combination of these principles makes that Nazarov has a GD of 27 out 37 as to you and others much less, but only because Nazarov belongs to a different line of R-L23 and others to your same line.
I wrote about an Indian R1a1a, then belonging to a recent subclade as to Western European R1a-M420, but with so many mutations as to the same Western European subclades and calculated its age with the current method obtaining a time similar to the infamous Zhivotovsky rate.
For this the calculation of the other infamous Anatole Klyosov based upon the 22 slowest markers merits attention and I have used it in my assessment of the R1 subclades. These slow mutating markers suffered less than the fastest ones these three principals, the mutations are above all of 1 plus or minus, and this demonstrates that there haven’t been mutations around the modal.
...

Any two haplotypes may contain anomalies and do not necessarily represent their subclades well so TMRCA estimates between just two people is subject to very large error ranges. However, a GD of 27 and 37 surely means we have only a very distant relationship. We do see instances of convergence between people of different haplogroups so it is possible to have a closer GD with someone in another haplogroup than within your own.

I think the application of examples like R1a-M420, discussions on mutation rates, etc. are generic topics can could distract from this thread focused on "New study on R1b Ht35 published by Lucotte." For that reason I quoted Rathna over here and we can go in-depth over there on Rathna's principles.
http://www.anthrogenica.com/showthread.php?828-STR-Wars-GDs-TMRCA-estimates-Variance-Mutation-Rates-amp-SNP-counting

TigerMW
08-02-2013, 10:48 PM
Maciamo Hay created a series of haplogroup maps. He recently put together this one. Rather than using R1b-L23xL51 he pretty much made it all of L23+ except P312+ and U106+. That takes out the big western subclades of R1b so I guess this is another way to look at things. I guess you could think of it as a shadow or perhaps a trail.
https://dl.dropboxusercontent.com/u/17907527/R1b-L23xP312x106_Frequency_Map_by_Eupedia.gif

http://www.eupedia.com/europe/maps_Y-DNA_haplogroups.shtml#R1b-L23

DMXX
08-02-2013, 11:19 PM
Which population is represented by the ht35 black hole in SE Anatolia? The Kurds?

Joe B
08-02-2013, 11:28 PM
Looked through the Assyrian, Aramaic, Armenian and Turkey projects. Did not see much sampling around Diyarbakır from the maps.
Maciamo Hay makes some very nice looking maps.

R.Rocca
08-02-2013, 11:52 PM
Maciamo Hay created a series of haplogroup maps. He recently put together this one. Rather than using R1b-L23xL51 he pretty much made it all of L23+ except P312+ and U106+. That takes out the big western subclades of R1b so I guess this is another way to look at things. I guess you could think of it as a shadow or perhaps a trail.


Unfortunately he kept L51 which we know is not part of Z2103/Z2105 which is what this map likely represents. It definitely would make a difference in a place like the Austrian/Italian border where L51(xL11) is the third most frequent group after U106 and U152. I'm sure it is also affecting the rest of north-central Italy and France as well. Even more puzzling he added L11 as well which is not ht35.

alan
08-03-2013, 11:25 AM
Unfortunately he kept L51 which we know is not part of Z2103/Z2105 which is what this map likely represents. It definitely would make a difference in a place like the Austrian/Italian border where L51(xL11) is the third most frequent group after U106 and U152. I'm sure it is also affecting the rest of north-central Italy and France as well. Even more puzzling he added L11 as well which is not ht35.

He needs to rename his map. Still, its an interesting map of L23 before the two big clades. I am not sure why he didnt just take it one step up higher and also include M269*. That would give a good impression of M269 before the two clades. Seems odd to have left it out to me.

Sometimes when I look at a map like that, nice though it is, you really have got to wonder if you can imply much by frequency given all that has happened in the last 5000 years. For example areas which had a large Slavic expansion are going to have had their frequencies massively reduced as was shown in the Tyrol study. The Balkans shows a similar affect and the massive drop in R1b basically corresponds to the Slavic languages areas of stength in the Balkans. The population in the western steppes largely dates to only the last few centuries. The whole Med. saw so many colonies and empires in succession. In general eastern Europe and SW Asia have had very complex histories. So I tend to look at these maps just to located above-noise presence rather than think they are likely to give any sort of real idea of frequency in the deep past.

Michał
08-03-2013, 12:41 PM
Which population is represented by the ht35 black hole in SE Anatolia? The Kurds?
This is of course very likely. Also, a similar hole in NW Balkans may correspond to the substantial Slavic input.

Generally, this distribution is nearly perfectly consistent with the hypothetical association of R1b-Z2105 with the "Graeco-Armenian-Thracoid" linguistic group (or "post-Western-Yamna" population in my favourite scenario), especially if the small Austrian/Italian concentration indeed corresponds to L51* (as suspected by Richard). All this fits very well the ancient Greek colonization pattern in the Mediterranean and Pontic regions. It is also worth noting that Albanian is considered by many linguists (including those ones who I respect the most) to be the only surviving ancestor of the Dacian language.

newtoboard
08-03-2013, 02:06 PM
Which population is represented by the ht35 black hole in SE Anatolia? The Kurds?

Maybe Zazaki speakers? I thought ht35 was well represented in Kurmanji speakers as you can see in regions immediately to the east and south of that ht35 black hole. It is interesting though because it makes the claims of Zazaki Kurds having significant Armenian ancestry look wrong.



This is of course very likely. Also, a similar hole in NW Balkans may correspond to the substantial Slavic input.

Generally, this distribution is nearly perfectly consistent with the hypothetical association of R1b-Z2105 with the "Graeco-Armenian-Thracoid" linguistic group (or "post-Western-Yamna" population in my favourite scenario), especially if the small Austrian/Italian concentration indeed corresponds to L51* (as suspected by Richard). All this fits very well the ancient Greek colonization pattern in the Mediterranean and Pontic regions. It is also worth noting that Albanian is considered by many linguists (including those ones who I respect the most) to be the only surviving ancestor of the Dacian language.


Seems pretty light in mainland Greece, Western Turkey, Cyprus and Crete for it to have that association imo. Why would this lineage be better represented in Armenians (and possibly Laz speakers)? Could you post some more information about that "Graeco-Armenian-Thracoid" linguistic group? I thought Thracian was a Satem language and Albanian is related to Illyrian not Dacian.

alan
08-03-2013, 02:40 PM
This is of course very likely. Also, a similar hole in NW Balkans may correspond to the substantial Slavic input.

Generally, this distribution is nearly perfectly consistant with the hypothetical association of R1b-Z2105 with the "Graeco-Armenian-Thracoid" linguistic group (or "post-Western-Yamna" population in my favourite scenario), especially if the small Austrian/Italian concentration indeed corresponds to L51* (as suspected by Richard). All this fits very well the ancient Greek colonization pattern in the Mediterranean and Pontic regions. It is also worth noting that Albanian is considered by many linguists (including those ones who I respect the most) to be the only surviving ancestor of the Dacian language.

I have read up a fair bit on the possbilities for Albanian and I have to agree that the displaced Dacians idea seems by far the most convincing. There is no question that they have been displaced from somewhere else more inland to the north and east as its clearly written into their lack of maritime vocab and pattern of loans which doesnt fit their current location near Greece. That is one reason why the concentration of M269 among Kosovars interest me. I think it probably moved with the Albanians from Dacia and has a displaced distribution.

In generally it looks like language zones, even dead ones, correspond to DNA patterns better than geography. Language barriers, particularly with Slavs seems to be pretty strong genetic ones too. We see the same with Germanic. Both are surely the impact of full scale folk movements that displaced a very large proportion of the pre-existing ones. In fact I tend to agree now after years in this hobby that elite dominance most often does not lead to language replacement i.e Normans in England, Visigoths in Spain, Franks in France and many other cases, even if there is a 5-10% input basically a new upper class. I dont think the reverse situation of the Roman empire imposing Latin really is relevant for most prehistoric conditions in Europe.

It does raise the question of R1b though because it is basically some sort of elite expansion model involving beaker people and their immediate predecessors that is being proposed rather than a folk movement. It does again raise the possibility that L11 is older than we think and therefore was old enough to have been encorporated into the Neolithic movements, perhaps not the very first waves into Europe but maybe those later ones associated with developed dairy pastoralism (which spread from the Bosphorus across most of temperate Europe c. 5200-4000BC) like Lengyel, TRB and various other areas where the first farmers were a little later in arriving. That sort of model would make it a folk movement.

It would of course raise all sorts of questions about the steppe model. It is not the same as a first farmers model but a link with middle Neolithic dairy pastoralists which would go back to north Anatolia with them impacting in various directions. If we see languages as usually relating to folk movements then that might be one of sorts. While the Uralic evidence for PIE now appears to be very very shaky we would still have to contend with copper age vocab in PIE, the wheel etc in such a model.

However, the copper age in SE Europe is early and older than farming in many parts of temperate Europe. The wheel has never totally concinced me as a vocab lynchpin (no pun intended). Non wheel vehicle terms for round or spinning things or bits of sleds etc could have been used. I have an open mind of the IE question as I never feel the evidence is quite as clinching as portrayed adn there is still plenty wriggle room IMO. I dont think its impossible that Anatolian was located in parts of Anatolia from the outset and that PIE evolved from middle Neolthic settlers from there who settled in the Balkans around 5200BC, early enough to be important in the rise of advance cattle pastoralism across a swath of European first Neolithic cultures in the north and other previously unsettled areas outside the thin LBK band and also as a secondary wave in areas already settled. It at least has the advantage of linking language change to something of a folk movement and seems compatible with R1b's phylogeny.

Palisto
08-03-2013, 04:21 PM
Which population is represented by the ht35 black hole in SE Anatolia? The Kurds?

I think the black hole is based on Kurmanji Kurds from Turkey in Nasidze et al., 2005.
R1*(x R1a1* M17): 4/87 (=4.6%).

The genetic publications about Kurds (or this specific region) show a different pattern:

The black hole also correspond to Region 4 (East Turkey) and Region 5 (Southeast Turkey) in Cinnioglu et al., 2004.

Region 4 (East Turkey):
P25 (xM269,M73,M335): 1/82 (1.2%)
M269: 10/82 (12.1%)
M73: 0/82 (0%)
M335: 0/82 (0%)

Region 5 (Southeast Turkey):
P25 (xM269,M73,M335): 0/43 (0%)
M269: 6/43 (14.9%)
M73: 0/43 (0%)
M335: 0/43 (0%)


From Iraqi Kurds in Stenersen et al., 2004 (I used the STR data to predict the haplogroup based on Athey's Haplogroup predictor.)
R1b: 13/107 (12.1%)

From Iranian Kurds in Grugni et al., 2012:
R1b* M343 1/59 (1.7%)
R1b1a1 M73 0/59 (0%)
R1b1a2* L23* 0/59 (0%)
R1b1a2a1a M412 0/59 (0%)

To summarize:
I don't think that there is a drop in L23 frequency in Kurdistan-Turkey/Kurmanji Kurds.

TigerMW
08-03-2013, 10:21 PM
I think the black hole is based on Kurmanji Kurds from Turkey in Nasidze et al., 2005.
R1*(x R1a1* M17): 4/87 (=4.6%)....

To summarize:
I don't think that there is a drop in L23 frequency in Kurdistan-Turkey/Kurmanji Kurds.

Do you mind if I quote you over on Eupedia to see what Maciamo's responses are? From what I've seen, he'll update his maps as new information or corrections come in.

Michał
08-03-2013, 10:53 PM
Seems pretty light in mainland Greece, Western Turkey, Cyprus and Crete for it to have that association imo.
The mainland Greece, including Thessalia and Macedonia, was much strongly affected by the Slavic migrations than Peloponesse. As for Cyprus and Crete, they were both previously inhabited by the highly civilized non-Greek populations (Minoans and Eteocypriots), so I would expect a much stronger contribution of some local non-R1b substratum. As for Western Turkey, I am not sure whether we should expect a much higher frequency of R1b-Z2105 on the coast than in Central Anatolia, considering a significant presence of some non-Greek (but closely related) populations in that part of Asia Minor, including for example the Phrygians.




Why would this lineage be better represented in Armenians (and possibly Laz speakers)?
There are many possibilities, including the bottleneck effect and some local differences related to the contribution of non-IE substratum. Armenians are considered to be related to Greeks, so I don't know why you find this relatively high frequency of R1a-Z2105 in Armenia unexpected. Anyway, do you know any better explanation for the significant R1b-Z2105 presence in Armenia (as well as in Greece, Southern Italy, Albania and Romania)?




Could you post some more information about that "Graeco-Armenian-Thracoid" linguistic group? I thought Thracian was a Satem language and Albanian is related to Illyrian not Dacian.
Alan has already provided some information on this subject, but you may also find some references in Wikipedia:
http://en.wikipedia.org/wiki/Dacian_language#Albanian
http://en.wikipedia.org/wiki/Origin_of_the_Albanians#Thracian_or_Dacian_origin

There were also numerous discussions on this subject held in the IE linguistic Yahoo group (cybalist):
http://tech.groups.yahoo.com/group/cybalist/message/22462
http://tech.groups.yahoo.com/group/cybalist/message/15112
http://tech.groups.yahoo.com/group/cybalist/message/15552

The Illyrian origin of the Albanian language is still supported by many (mostly Albanian) linguists, but the hypothetical Dacian-Albanian relationship seems to be gaining more supporters recently. As a matter of fact, some people consider Illlyrian to be related to Daco-Thracian, which would make Albanian a member of this entire group irrespective of whether it originates from Dacian or Illyrian.

As for the Satem issue, my hypothesis assumes that the ancestors of ancient Greeks were represented by the westernmost part of Yamna (Ushatovo?), so their contacts with the Eastern Yamna (Satem-speaking) members of R1a-Z93 were not that strong and they were thus able to avoid Satemization unlike their more Eastern relatives (i.e. the hypothetical ancestors of Daco-Thracians).

Silesian
08-03-2013, 11:12 PM
Do you mind if I quote you over on Eupedia to see what Maciamo's responses are? From what I've seen, he'll update his maps as new information or corrections come in.

Perhaps he is trying to show the gap where he places the Mitanni as primary R1a.

http://www.eupedia.com/europe/Haplogroup_R1a_Y-DNA.shtml

However this scenario does not show the full picture as R1b -L23 flanks in Çorum https://en.wikipedia.org/wiki/%C3%87 [ Hittite heartland and to the Northwest]
Armenian Syunik-Karabakh Provinces to the North East
Gilaki/Talysh/Lur/ to the East and South East; not shown in his map.


Quote Originally Posted by newtoboard View Post
Seems pretty light in mainland Greece, Western Turkey, Cyprus and Crete for it to have that association imo.

https://en.wikipedia.org/wiki/Bryges

https://en.wikipedia.org/wiki/File:Map_of_ancient_Epirus_and_environs.png


The evolution of Proto-Greek should be considered with the background of an early Palaeo-Balkan sprachbund that makes it difficult to delineate exact boundaries between individual languages. The characteristically Greek representation of word-initial laryngeals by prothetic vowels is shared by the Armenian language, which also shares other phonological and morphological peculiarities of Greek. The close relatedness of Armenian and Greek sheds light on the paraphyletic nature of the Centum-Satem isogloss.

https://en.wikipedia.org/wiki/Proto-Greek_language

https://en.wikipedia.org/wiki/File:Proto_Greek_Area_reconstruction.png

https://en.wikipedia.org/wiki/Sicani

Some modern scholars think the Sicani may have been an Illyrian tribe.....Pottery found by archaeologists at Skorba resembles that found in Italy, and suggests that the Maltese islands were first settled in 5200 BC mainly by stone age hunters or farmers who had arrived from the larger island of Sicily, possibly the Sicani

DMXX
08-04-2013, 04:05 AM
[B]I think the black hole is based on Kurmanji Kurds from Turkey in Nasidze et al., 2005.
...

To summarize:
I don't think that there is a drop in L23 frequency in Kurdistan-Turkey/Kurmanji Kurds.

If the data for Kurmanji-speaking Kurds in Turkey is from Nasidze et al., the author of these maps isn't well-acquainted with the numerous irregularities that exist in Nasidze's results concerning West Asia.

In data from what I'm presuming is the same paper, >20% Y-DNA I was reportedly found in Tehran, Iran. That result has never been replicated in any studies that followed. That isn't the only spurious result observed in that paper.

I would advise the map author to not include any results from that paper and instead opt for Cinnioglu's. Regardless of linguistic affiliation, the Iranian Kurds are not the same source population, so Grugni's numbers shouldn't even be considered anyway.

Palisto
08-04-2013, 06:54 AM
Do you mind if I quote you over on Eupedia to see what Maciamo's responses are? From what I've seen, he'll update his maps as new information or corrections come in.

No problem, let's see what he has to say.


If the data for Kurmanji-speaking Kurds in Turkey is from Nasidze et al., the author of these maps isn't well-acquainted with the numerous irregularities that exist in Nasidze's results concerning West Asia.

In data from what I'm presuming is the same paper, >20% Y-DNA I was reportedly found in Tehran, Iran. That result has never been replicated in any studies that followed. That isn't the only spurious result observed in that paper.

I would advise the map author to not include any results from that paper and instead opt for Cinnioglu's.
I agree.



Regardless of linguistic affiliation, the Iranian Kurds are not the same source population, so Grugni's numbers shouldn't even be considered anyway.
Data for Kurmanji Kurds should be based on Kurmanji Kurds, and Iranian Kurds are not Kurmanji Kurds. I just want to show the full picture for Kurds.

alan
08-04-2013, 12:13 PM
i can see another issue with this. He doesnt use this study on Moldovians http://forwhattheywereweare.blogspot.co.uk/2013/01/y-dna-of-moldovans.html

This uses weird SNP nomenclature but basically tested for M269, L51, U106 and M73. It gives at least a total for M269xL51 which is likely mostly L23. This averages over 7% but reached nearly 10% among the north Moldovan sample and the Ukrainian speakers in the east of the country. The average was dragged down by a signidicantly lower total from a village in the extreme SE of Moldova. I suspect an average of just short of 10% may be most representative. Regardless of the detail the map appears to be wrong in that Moldova should all be in the next shading band up.

T101
08-04-2013, 03:52 PM
All this fits very well the ancient Greek colonization pattern in the Mediterranean and Pontic regions. It is also worth noting that Albanian is considered by many linguists (including those ones who I respect the most) to be the only surviving ancestor of the Dacian language.

It is nice when when these suspected pieces start fitting together so nicely. I wish though they had tested cities in Southern France with a Greek foundation rather than Montpellier and Grasse. The absence in Southern France does give some pause, but a remarkable fit nonetheless given all the tumultous history of the Mediterranean region.





As for the Satem issue, my hypothesis assumes that the ancestors of ancient Greeks were represented by the westernmost part of Yamna (Ushatovo?), so their contacts with the Eastern Yamna (Satem-speaking) members of R1a-Z93 were not that strong and they were thus able to avoid Satemization unlike their more Eastern relatives (i.e. the hypothetical ancestors of Daco-Thracians).

The Dacian connection looks solid too. All the Carpathian hot spots such as Slovakia (6%), Hungary (7.4%), Romania (10.3%) are represented well. The Satemisation however most likely came from the north with Z280 groups rather than from the east with Z93 groups (i.e. Armenians). There is a close parallel between Baltic placenames and Daco-Thracian ones. Dacian could therefore possibly share an ancestor with the modern Baltic languages. As the Proto-Baltic-Slavic Urheimat (Trzciniec culture) bordered the Eastern Carpathians.

Silesian
08-05-2013, 04:44 PM
i can see another issue with this. He doesnt use this study on Moldovians http://forwhattheywereweare.blogspot.co.uk/2013/01/y-dna-of-moldovans.html

This uses weird SNP nomenclature but basically tested for M269, L51, U106 and M73. It gives at least a total for M269xL51 which is likely mostly L23. This averages over 7% but reached nearly 10% among the north Moldovan sample and the Ukrainian speakers in the east of the country. The average was dragged down by a signidicantly lower total from a village in the extreme SE of Moldova. I suspect an average of just short of 10% may be most representative. Regardless of the detail the map appears to be wrong in that Moldova should all be in the next shading band up.

Dug up some old posts from Palamede, on other forums. More edits for the map as per the following

http://www.ncbi.nlm.nih.gov/pmc/articles/PMC3068964/


The phylogenetic relationships and haplogroup frequencies for the data from the two sites in Asia Minor: Phokaia and Smyrna, three mainland Greek sites, the four regions from Turkey and the Neolithic sites in Provence are given in Figure ​Figure2.2. Phokaia and Smyrna have just subtle differences in their haplogroup composition. The dominant haplogroups in both Phokaia and Smyrna are E-V13 (19.4% and 12.1%) and R1b-M269 (22.6% and 27.8%) respectfully. In addition, J2a is also common, attaining a frequency of 9.7% in Phokaia and 15.5% in Smyrna. This high frequency of haplogroup J2a-Page55 (formerly DYS413≤ 18) in Smyrna is characteristic of non-Greek Anatolia........The dominant haplogroup of Provence is R1b-M269 at 58.8% (Figure ​(Figure2).2). Also found in Provence is haplogroup E-V13 (3.9%) and J2a-DYS445 = 6 (7.8%).

Only a small percentage Provence are perhaps L23x51.

Smyrnia settlements start to overlap with also Luwian Language areas and it's importance with regards wheel.
https://upload.wikimedia.org/wikipedia/commons/3/34/Luwian_Language_de.svg

Luwian was among the languages spoken during the 2nd and 1st millennia BC by groups in central and western Anatolia and northern Syria.[4] In the Old Hittite version of the Hittite Code,.....Luwian has also been enlisted for its verb kalut(t)i(ya)-, which means "make the rounds of" and is probably derived from *kalutta/i- "circle".[19] It has been argued[20] that this derives from a proto-Anatolian word for "wheel", which in turn would have derived from the common word for "wheel" found in all other Indo-European families. The wheel was invented in the 5th millennium BC and, if kaluti does derive from it, then the Anatolian branch left PIE after its invention (so validating the Kurgan hypothesis as applicable to Anatolian). However, kaluti need not imply a concrete wheel and so need not have been derived from a PIE word with that meaning. The IE words for a wheel may well have arisen in those other IE languages after the Anatolian split.







R1b 16,9% north-south cline: Macedoinia-Thracia 13,2% Epiria-Thessalia 10,2% Peloponnese 20,5%, Crete 17,1% (15,3%), Egean islands and the refugees from Anatolia,22,8% including Chios 26,2%, Fokia (Ionia) 22,6% Smyrnia (Eolia) 27,6%

R1b1* Greece 0% Crete 0,5%
R1b1a2*-M269 Greece 2% Crete 3%

Of interest is also the two regions listed above may have contributed to the Philistines colonizing the 5 cities in Levant.


There is some limited evidence in favour of the assumption that the Philistines were Indo-European-speakers either from Greece and/or Luwian speakers from the coast of Asia Minor. Philistine-related words found in the Bible are not Semitic, and can in some cases, with reservations, be traced back to Proto-Indo-European roots.[5] By the beginning of the 1st Millennium BC they had adopted the general Canaanite language of the region.

https://www.youtube.com/watch?v=_PYGoGfyx6k

2:32 part two possible homeland.



Proportion of R1b in Near and Middle East, Anatolia and South Balkans
Ararat Valley 37,3% Gardmann 31,3% Lake Van 32% Sasun 15,4% W Turkey 13,5% Centre Turkey 19,1% East Turkey 12% Syria 3,7% Jordania 9,2% Liban 7,6% Iran 8,6% Caucasus 3,0% Greece 13% Macedonia 12,8%
Phokaia (31) 23% Smyrnia (58) 28% NeoNikomedia (57) 19% Sesklo/Dimini(57) 5% Lerna/Franchti(57) 16% NW Anatolia(52) 12% Med Anatolia(33) 21% Central Anatolia (90) 16% W Anatolia(30) 17%

Silesian
01-02-2016, 06:51 PM
Dug up some old posts from Palamede, on other forums. More edits for the map as per the following

http://www.ncbi.nlm.nih.gov/pmc/articles/PMC3068964/



Only a small percentage Provence are perhaps L23x51.

Smyrnia settlements start to overlap with also Luwian Language areas and it's importance with regards wheel.
https://upload.wikimedia.org/wikipedia/commons/3/34/Luwian_Language_de.svg








Of interest is also the two regions listed above may have contributed to the Philistines colonizing the 5 cities in Levant.



https://www.youtube.com/watch?v=_PYGoGfyx6k

2:32 part two possible homeland.

Jan 2 2016 autosomal update/study for above diagram/general region.
To date no R1b found in ancient samples.
http://www.cell.com/current-biology/abstract/S0960-9822%2815%2901516-X

Barellalee
04-22-2016, 05:26 AM
Very interesting read. Autosomal Matches to the Surname Sarni from Morra De Sanctis, Italy, which was my GGGrandmother's surname and town, have shown that her Father was R1b L23. The Sarnis had been in Morra De Sanctis, which is in the Avellino Province of Campania, for centuries. Going back to the early Medieval Period, the origin of this line was the town of Sarno, at the base of Mount Vesuvius, in the Salerno Province. So our ancestral origin was in Campania, the region immediately North of where L23 peaks, in Calabria. The Ancient Greek theory is interesting, but from all the information I know, including all of the places this Haplogroup has shown up in Italy, it appears far too common and widespread to be attributed to Greeks in Italy, at least all of it. Arbereshe are showing in the Balkan cluster of the group and I imagine that's expected. Our Sarni line is not of Arbereshe origin, and Autosomal Tests, both 23andMe and dna land, have so far shown 0% Balkan Autosomal contribution in my Family, which is Tuscan, Abruzzese, and Campanian.

Silesian
02-25-2017, 03:03 PM
Interesting place Italy.

Here is one who matches me at 12, from Schleswig-Holstein. Exact at 12 but extended we differ, mostly on the fast markers, which is why he belongs to 7.0 [L21+] P312+ M222- Extended and chosen from 37.
8/8 slow marker match[mutation rate]

393-12-0.00076%
426-12-0.00008%
388-12-0.00022%
392-13-0.00052%
455-11-0.00016%
454-11-0.00016%
437-15-0.00099%
438-12-0.00055%

13/17 medium match [mutation rate]

390-24-0.0031%
19-14-0.00151
391-10-0.00226%
385-11-14-0.00226%
439-11-0.00477%[/I]
389I-14-0.00186%
389II-31-0.00242%
459-9-10-0.00132%
447-25-0.00264%
448-19-0.00135%
YCAII-19-23-0.00151%-*DYS394
3 markers 1 step.
460-11-10
Gata-H4-11-10
442-11-12

2/10 fast match
464-b-c-15-16-0.00566%

Update-R1b-M269>L23>Z2103> found in Cradle of Kurgan Culture.
Waiting for new results to see if predicted sample P312 sample[from Schleswig-Holstein] matched at first string of [12]strs and above combination of slow and fast markers are connected with Cradle of Kurgan Culture[5000ybp+/-] or not[ as speculated almost 4 years ago:)] .

Silesian
05-27-2017, 11:25 AM
May 26 2017 update; Again as above study no R1b found. However interesting Steppe results.



http://pichoster.net/images/2017/05/26/a8c11f94b5eb80e6e2c2f343e39b2dcc.jpg


Continuity and admixture in the last five millennia of Levantine history from ancient Canaanite and present-day Lebanese genome sequences
http://biorxiv.org/content/early/2017/05/26/142448

Mis
06-01-2017, 08:04 PM
delete