World Families Forums

General Forums - Note: You must Be Logged In to post. Anyone can browse. => R1b General => Topic started by: polako on April 30, 2010, 01:46:19 PM



Title: A Comparison of Y-Chromosome Variation in Sardinia and Anatolia
Post by: polako on April 30, 2010, 01:46:19 PM
Quote
A Comparison of Y-Chromosome Variation in Sardinia and Anatolia Is More Consistent with Cultural Rather than Demic Diffusion of Agriculture

Individually, R-M269 represents the most informative lineage for which previous studies have reached conflicting conclusions. Therefore, to have a more definitive picture we constructed a network of R-M269 considering not only Anatolia and Sardinia but also the Balkan, Georgian, Iberian and North-Western European populations (figure 2A).

A dichotomy between Western and Eastern populations was apparent with two distinct core haplotypes, corresponding to two informative R-M269 STR patterns. On the one hand is the DYS393-13/DYSA7.2-12 STR pattern common throughout Western Europe and the Iberian peninsula, the Atlantic Modal Haplotype [16]. On the other hand is the DYS393-12/DYSA7.2-11 STR pattern which appears as a more recent Eastern European haplotype.

The Sardinian haplotypes belong to the Atlantic Modal Haplotype variability, with an interesting internal differentiation shown by the completely Sardinian branch off-shoot (figure 2A). In contrast, the majority of Anatolian samples belong to the DYS393-12/DYSA7.2-11 subtype. Interestingly, the bridge between the two main forms, is represented by the intermediate step of a haplotype common in the Balkan region, DYS393-13/DYSA7.2-11. This dichotomy is further corroborated when TMRCA values for R-M269 are examined; they provided a value of 32.6 KYA (1000 Years Ago, C.I. 95% 25.0–80.7) in the Iberian sample, 27.0 KYA in the Sardinians, (C.I. 95% 19.5–67.5) and 19.6 KYA (C.I. 95% 19.4–44.4) in the Anatolians: in all cases clearly pre-dating the advent of agriculture.

Discuss...

http://www.plosone.org/article/info%3Adoi%2F10.1371%2Fjournal.pone.0010419


Title: Re: A Comparison of Y-Chromosome Variation in Sardinia and Anatolia
Post by: Mike Walsh on April 30, 2010, 01:55:49 PM

I don't know if this is worth discussing.

There is a huge conflict between the dates.
Quote
MRCA values for R-M269 are examined; they provided a value of 32.6 KYA (1000 Years Ago, C.I. 95% 25.0–80.7) in the Iberian sample, 27.0 KYA in the Sardinians

Per the 2008 study, with Behar, Hammer, et al...  R1, itself, has a TMRCA of only 18.5K ybp.  R-M269 can't be anywhere near as old as what this article is claiming.


Title: Re: A Comparison of Y-Chromosome Variation in Sardinia and Anatolia
Post by: polako on April 30, 2010, 02:04:11 PM
Per the 2008 study, with Behar, Hammer, et al...  R1, itself, has a TMRCA of only 18.5K ybp.  R-M269 can't be anywhere near as old as what this article is claiming.

Yeah, it's a pretty crazy article in parts. They claim that E, J and G in Europe are all Paleolithic, and the 36KYA estimate for M269 is really out there, that's true.

However, what about the marker they use, the DYSA7.2? Is that as useful as they say?



Title: Re: A Comparison of Y-Chromosome Variation in Sardinia and Anatolia
Post by: vtilroe on April 30, 2010, 02:38:29 PM
Per the 2008 study, with Behar, Hammer, et al...  R1, itself, has a TMRCA of only 18.5K ybp.  R-M269 can't be anywhere near as old as what this article is claiming.

Yeah, it's a pretty crazy article in parts. They claim that E, J and G in Europe are all Paleolithic, and the 36KYA estimate for M269 is really out there, that's true.

However, what about the marker they use, the DYSA7.2? Is that as useful as they say?



Read through the Discussion section.  I tend to agree with Mikewww - this is hardly worth paying attention to.  Bikini haplotypes, Zhiv rates... c'mon.  Some people never learn.


Title: Re: A Comparison of Y-Chromosome Variation in Sardinia and Anatolia
Post by: Maliclavelli on April 30, 2010, 04:29:15 PM
Friends, even though you divide for 3 the Zhiv. rate nothing changes: also Anatolian TMRCA will be divided for 3 and we would be around the Younger Dryas anyway.
This paper is in line with what I was being saying from many years. The last analyses done also by Argiedude are even more definitive: we have the path of Western European (Italian/Sardinian) hg. R, and Italy has certainly the presence of R1b1*, R1b1a (M18), till R-L51 (the highest percentage in the world), and the last findings (see other topics here) also of R-P132 etc.


Title: Re: A Comparison of Y-Chromosome Variation in Sardinia and Anatolia
Post by: polako on April 30, 2010, 09:03:34 PM
I can't get past that 36KYA. If they made a mistake like that, then I'm not sure how reliable the rest of the article is.

But what I was wondering about was whether the data in this travesty of a report maybe supported an early post-Ice Age expansion of R1b from west Asia to Europe, and then diversification within Europe.

This would fit my theory that R1b wasn't linked to the early Indo-Europeans at all. It only become Indo-European during the Iron Age, and exploded with the Celts, and then the Franks etc. On the other hand, in Asia, it basically became a Turkic haplogroup.


Title: Re: A Comparison of Y-Chromosome Variation in Sardinia and Anatolia
Post by: Maliclavelli on May 01, 2010, 02:34:07 AM
The problem isn't which method of calculation they used but if it is reliable their variance. If you divide for 3 the 32.6YBP of Iberians, the 27.0 of Sardinians and the 19.6 of Anatolians, Iberians and Sardinians aren't derived from Anatolians.

About the method I have always said that, given the mutations around the modal, perhaps Zhivotovski is too ancient but also to divide for 3 his calculation isn't reliable.


Title: Re: A Comparison of Y-Chromosome Variation in Sardinia and Anatolia
Post by: Maliclavelli on May 01, 2010, 06:28:01 AM
Argiedude, probably Contu has replied to our question about that R-M18: perhaps now it is among R-M269: CA244. But it is very interesting, having this cluster a very high variance: see CA236, CA4 (4 samples) etc.

It seems a cluster like: 14, 23, 11, 14, 13, 13-29 (but with many samples from 13 to 15 and with mutations also in the second marker), 12, 13, 12. Presupposing the mutations around the modal, this cluster is very very ancient.


Title: Re: A Comparison of Y-Chromosome Variation in Sardinia and Anatolia
Post by: Maliclavelli on May 01, 2010, 06:53:23 AM
Diffused in the West Mediterranean Sea (and among Mestizos), there are on YHRD a sample from Majorca with DYS389=15-31) and with DYS437=14. We don't know if Sardinians come from Spain or Spaniards from Italy (but Contu presupposes a more ancientness of the Spaniards like the old Cantabrian Refugium hypothesis). It could be an R1b1b2a1b.


Title: Re: A Comparison of Y-Chromosome Variation in Sardinia and Anatolia
Post by: Maliclavelli on May 01, 2010, 06:59:17 AM
Given its distribution, it could be a very early  "Out of Italy" of R-P312.


Title: Re: A Comparison of Y-Chromosome Variation in Sardinia and Anatolia
Post by: vineviz on May 01, 2010, 08:30:46 AM
However, what about the marker they use, the DYSA7.2? Is that as useful as they say?

It's another name for DYS461, and it can be helpful in distinguishing some of the basal paragroups from the downstream clades.

But Morelli et al. lacked the SNPs to create a sub-M269 phylogeny, they misinterpret the polarity of DYS461 and draw the wrong conclusion from its use.

Another laughable paper from Morelli and Contu.  What a shame.


Title: Re: A Comparison of Y-Chromosome Variation in Sardinia and Anatolia
Post by: Maliclavelli on May 01, 2010, 11:13:35 AM
It seems that two samples from Sorgono with DYS393=12, have DYS461=10 and also DYS385=11-16. This is probably an R1b1b2a (L23+), with more variance of the Italian and the Anatolian ones. I'll try for this cluster on YHRD.


Title: Re: A Comparison of Y-Chromosome Variation in Sardinia and Anatolia
Post by: Maliclavelli on May 01, 2010, 11:22:09 AM
This cluster is present in a few cases: my Tuscany, Hungary, France and Western Europe. Then, dear Vizachero, I confirm all my hypotheses: the most ancient R1b1b2/L23+ is here and perhaps in Sardinia or in these places you should find for L150-, not where you are finding thus far.


Title: Re: A Comparison of Y-Chromosome Variation in Sardinia and Anatolia
Post by: argiedude on May 01, 2010, 12:55:31 PM
"an interesting internal differentiation shown by the completely Sardinian branch off-shoot"

http://www.plosone.org/article/slideshow.action?uri=info:doi/10.1371/journal.pone.0010419&imageURI=info:doi/10.1371/journal.pone.0010419.g002

I looked at the STR data and yep, there's a definite cluster, which became readily apparent thanks to the extra STRs in this new study. It has 390=23, 439=13, 389a=14, 392=14. It makes up about 1/6 of Sardinia's R1b1b2.

.............................

In my variance map of ht15 I had a value of 0,21 for Sardinia, with 15 samples. Now, with 56 samples, I redid the estimate and I got... 0,21. Morelli got a higher value because she included the samples from the cluster I noted above. I always exclude any clusters that have a different modal haplotype than the general R1b1b2 modal. Redoing the calculation, this time including those samples, I get around 0,27, same as Morelli's estimate.

.............................

"the STRs used in our study are more numerous than in the work of Balaresque et al.; specifically we typed also the marker DYS461 and they did not. This marker is critical for haplotype identification. In fact, when we repeated our analysis but excluded marker DYS461, the resulting network goes from a markedly bipolar structure to a starlike one that strongly resembles that of Balaresque et al."

.............................

A few days ago I noted that there was a North African cluster of R1b1b2, with 385b=13.2, amongst other values, such as DYS19=13 and 389a=14. Contu's samples, but not Morelli's, have an R1b1b2 sample with 385b=13.2. The rest of its haplotype is different from the North African cluster. But it has 389a=14. I had found several European R1b1b2 with 385b=13.2, and they mostly had a typical R1b1b2 haplotype, except all of them shared 389a=14 (one of the modal values of the North African cluster). That's probably not a coincidence, so we're looking at 2 clusters of R1b1b2 with 385b=13.2, one from Europe and one from North Africa. This Sardinian sample is yet another member of the European group. Both groups must obviously be related. Too bad Morelli didn't test one of these samples. We have to find some client of FTDNA who belongs to this haplogroup and have him fully tested. The North African cluster makes up 1% of their y-dna. With just 100 clients, there should be one that belongs to this cluster. By the way, one North African sample seemed to belong to the European cluster, and 2 European samples, both from Iberia, clearly belonged to the North African cluster.

.............................

Argiedude, probably Contu has replied to our question about that R-M18: perhaps now it is among R-M269: CA244. But it is very interesting, having this cluster a very high variance: see CA236, CA4 (4 samples) etc.

I didn't understand anything you said. You're talking about M18 but you mention samples that are all M269?

It seems a cluster like: 14, 23, 11, 14, 13, 13-29 (but with many samples from 13 to 15 and with mutations also in the second marker), 12, 13, 12. Presupposing the mutations around the modal, this cluster is very very ancient.

A cluster of what? M269? M18?

However, what about the marker they use, the DYSA7.2? Is that as useful as they say?

It's another name for DYS461, and it can be helpful in distinguishing some of the basal paragroups from the downstream clades.

That's an understatement, it's more like almost perfect in distinguishing ht35 from ht15.

But Morelli et al. lacked the SNPs to create a sub-M269 phylogeny,

The haplotype data they used is almost perfect for separating the 2 groups. There only seemed to be a problem with the Balkan samples, where they found that half of them were "intermediate", having 393=13 and 461=11, but we now know that these Balkan samples are in fact an Albanian cluster of ht35, which also have 385=11/11, and they're a very recent clade, so they blew it on that one particular group, but other than that, 393=12 + 461=11 is equivalent to ht35 with extremely few exceptions, and 393=13 + 461=12 is equivalent to ht15 with very few exceptions, so their analysis isn't flawed in this regard. Ken Nordtvedt has said that he prefers haplotype data to SNP testing.

they misinterpret the polarity of DYS461 and draw the wrong conclusion from its use.

What does that mean?

Another laughable paper from Morelli and Contu.  What a shame.

And what were the previous laughable studies from them?


Title: Re: A Comparison of Y-Chromosome Variation in Sardinia and Anatolia
Post by: Maliclavelli on May 01, 2010, 02:18:32 PM
Certainly Vizachero was referring to the paper of Contu et alii "Y-Chromosome Based Evidence for Pre-Neolithic Origin..." "PLos ONE" 2008 where the authors were the same in a different order and where was that R-M18 about which we discussed (you, Vizachero and me) and about which I wrote to Contu: Cagliari: 14,11-14, 13-29, 23, 11, 13. You said it wasn't R-M18 but R-M269, and now probably it has been tested again and it is really R-M269.
I confide in my memory and think that others do the same.

Nothing changed about the rest.


Title: Re: A Comparison of Y-Chromosome Variation in Sardinia and Anatolia
Post by: Maliclavelli on May 01, 2010, 02:29:58 PM
Argiedude writes: "A cluster of what? M269? M18?"

If you look at the paper this cluster is clearly tested M269 as I have said to you about your "North African " cluster, that I think having demonstrated not only that it is M269 but also clearly of European (Sardinian? Italian?) origin.


Title: Re: A Comparison of Y-Chromosome Variation in Sardinia and Anatolia
Post by: Maliclavelli on May 02, 2010, 02:57:07 AM
It is worth noting that Contu 2008 has an haplotype like that from Brescia (DYS385b=13.2) I posted on the thread on the North African haplotype Argiedude spoke about on this forum, then it is strengthned my hypothesis that it is an Italian one and the African is derived. It can have come from Italy like the R1b1* which generated all the African R1b1 and subclades or in a second wave, being this probably an R1b1b2a1b.
Contu 2008 has also 4 R1* (M173) unfortunately not tested on Morelli 2010 and 4 R-M17 very different among them and I remember again my theory that also R1a can have come from the Italian Refugium. I signalled in the past some ancient R1a in the Rhaetian region with DYS392=13. Unfortunately Morelli 2010 hasn't tested them for this marker.


Title: Re: A Comparison of Y-Chromosome Variation in Sardinia and Anatolia
Post by: Maliclavelli on May 02, 2010, 03:26:38 AM
R1a (Contu 2008)
Haplotype 1 (14, 10-15, 13-31, 25, 10, 14) finds the closest to itself in Bialystok (Poland) with DYS19=13, DYS392=14, DYS438=10, DYS439=11

Haplotype 2 (16, 11-14, 14-31, 24, 10, 14) finds the closest to itself in East Europe, both with DYS392=11 or 13

Haplotype 3 (16, 11-14, 13-30, 25, 10, 13) is ubiquitous, but prevails in East Europe.

Then it would seem that Sardinian R1a-s have an Eastern European origin. Only a deep clade exam and above all the aDNA will answer if R1a has had origin in the Italian refugium from R1* found in Sardinia or not.


Title: Re: A Comparison of Y-Chromosome Variation in Sardinia and Anatolia
Post by: Maliclavelli on May 02, 2010, 04:52:44 AM
Re. Sardinian R1 (M173) I can say that the closest over the world have been found in Germany, Nepal, Portugal (I don't know the fourth because my research on YHRD is over for to-day). But on SMGF there are a preponderance of Western Europe and above all of Hiberia. I found a Brasilian Menezes I'll put on Ysearch with 43 markers. Then that R1* was in the Cantabrian Refugium before the LGM I think very likely.


Title: Re: A Comparison of Y-Chromosome Variation in Sardinia and Anatolia
Post by: Maliclavelli on May 02, 2010, 05:16:24 AM
There is the possibility that Menezes (CXD9V) is haplogroup T, but he matches closely also a tested R1b (Mathis: T4P4V). If the 4 Sardinians had been tested R-M173, then the closest to them could be hg. T but also hg. R1*.
Here SNPs are needed.


Title: Re: A Comparison of Y-Chromosome Variation in Sardinia and Anatolia
Post by: Maliclavelli on May 02, 2010, 05:29:22 AM
Athey's Y-Haplogroup predictor has no doubts it is hg. T. Then another flaw of Contu & Co.?


Title: Re: A Comparison of Y-Chromosome Variation in Sardinia and Anatolia
Post by: Maliclavelli on May 02, 2010, 05:42:45 AM
Now Mathis in the Haplogroup T project of FTDNA is under hg. T. Please, update Ysearch!


Title: Re: A Comparison of Y-Chromosome Variation in Sardinia and Anatolia
Post by: argiedude on May 02, 2010, 12:30:55 PM
I signalled in the past some ancient R1a in the Rhaetian region with DYS392=13. Unfortunately Morelli 2010 hasn't tested them for this marker.

Where did you see these? I looked at 2 studies of South Tyrol and didn't find R1a with 392=13.


Title: Re: A Comparison of Y-Chromosome Variation in Sardinia and Anatolia
Post by: argiedude on May 02, 2010, 12:39:49 PM
What this study did was pretty much what I did when I built those ht15 and ht35 variance maps. They realized R1b1b2 has 2 main modals (what we've all known for years, but most geneticists seem to be stuck in a time warp) and that this undermines everything about Balaresque's analysis, which erroneously claimed R1b1b2 has a start-like pattern, which we know for a fact isn't true because SNP-testing shows the vast majority of Anatolian R1b1b2 is ht35, while most European R1b1b2 is ht15. The Morelli study "found" this (the time warp, again) using the marker DYS461, and showed that Balaresque's star-pattern is wrong. They mention that study frequently, it would seem their study is a rebuttal to Balaresque's study.


Title: Re: A Comparison of Y-Chromosome Variation in Sardinia and Anatolia
Post by: Maliclavelli on May 02, 2010, 12:45:25 PM
The paper was that of Pichler. I quoted these haplotypes on this forum, perhaps on that on "The Italian Refugium". Now I have no possibility to search for them, but, as have said to you many times, I confide in my memory and am sure on this. When I am home I'll try for them, if you will have not found them before.


Title: Re: A Comparison of Y-Chromosome Variation in Sardinia and Anatolia
Post by: argiedude on May 02, 2010, 01:34:05 PM
The paper was that of Pichler. I quoted these haplotypes on this forum, perhaps on that on "The Italian Refugium". Now I have no possibility to search for them, but, as have said to you many times, I confide in my memory and am sure on this. When I am home I'll try for them, if you will have not found them before.

I found it. They have some 7 samples that are listed as R1a and they have 392=13, BUT... I seriously think this is simply an error, which we always have to keep in mind seems to be something very common in almost all y-dna studies. Everything else about their haplotypes indicates that they're R1b, not R1a. And the study also lists 5 other samples as K(xP), but their haplotype indicates that they are almost certainly R1b.


Title: Re: A Comparison of Y-Chromosome Variation in Sardinia and Anatolia
Post by: rms2 on May 02, 2010, 05:37:16 PM
What this study did was pretty much what I did when I built those ht15 and ht35 variance maps. They realized R1b1b2 has 2 main modals (what we've all known for years, but most geneticists seem to be stuck in a time warp) and that this undermines everything about Balaresque's analysis, which erroneously claimed R1b1b2 has a start-like pattern, which we know for a fact isn't true because SNP-testing shows the vast majority of Anatolian R1b1b2 is ht35, while most European R1b1b2 is ht15. The Morelli study "found" this (the time warp, again) using the marker DYS461, and showed that Balaresque's star-pattern is wrong. They mention that study frequently, it would seem their study is a rebuttal to Balaresque's study.

Except what is important about "HT35" (a term I wish we could quit using, since it is nearly useless) is that it is believed to be the ancestral form of R1b1b2. Thus the presence of a greater amount of P310- R1b1b2 would seem to be the mark of the older homelands of R1b1b2. Both so-called HT35 and so-called HT15 are M269+. They are not really two completely separate and distinct things, so I disagree with you about the usefulness of Balaresque et al.

Finding places that are almost exclusively P310+ ("HT15", another term we are better off without) doesn't tell us much about the origin of R1b1b2, although those with greater P310+ haplotype variance can tell us something about the origin and path of P310.


Title: Re: A Comparison of Y-Chromosome Variation in Sardinia and Anatolia
Post by: Mike Walsh on May 02, 2010, 10:11:30 PM
.... "HT35" (a term I wish we could quit using, since it is nearly useless) ....
Here! Here!  All, please consider using the proper and current terminology.  If you mean P310- or whatever please describe as such rather than using outdated (and obsolete) descriptors.  If you mean "pseudo(or proxy)" P310- versus confirmed P310- also be sure to be explicit on that as well.