Variance of R1b-L23+ Calculated from Academic Studies

[JeanL]

Here is the calculated variance of R1b-L23+(i.e. Everything downstream of R1b-L23 and R1b-L23(xL51)) by countries from the haplotype data found in Table-S3 of Myres et al.2010.



http://i1133.photobucket.com/albums/m582/jeanlohizun/Myresetal2010R1b-L23Variance.jpg

[Maliclavelli]

What should this demonstrate? The origin of R-L23? The origin of the European subclades, from R-L51 and downstream?
First of all R-L23* are only two, for what I know, i.e. who is also L150-: Romitti (Italy) and Seymour (Great Britain). We have the most part of R-L23 who are L150+. And of course from one of these was born R-L51+. Now, from the ancient map of Argiedude and me, and now from the last map of Richard Rocca, we can see that the centre of R-L51+ is Italy or Italy/France, then that ancestor R-L150+ probably should be found here and not elsewhere, and at this point your variance is worth nothing. We should search the haplotypes closest to R-L51 and subclades.
Your variance is due to many R-L23/L150+ which have nothing to do with our subclades, but which have cumulated variance just for being distant from our European subclades.
Magoon has posted on Rootsweb the new SNPs found in the 1KGP. Many are downstream L150

Z2103 to Z2110

Others are downstream L51

Z2111 to Z2119 and Z2120

Only from L11 descend the subclades P312 and U106 and with this SNP is HG00148.

It seems that Iberia and South America are plenty of R-L11.
Then search on West and not East, you who are born on the Pyrenees.

[Jdean]

Here is the calculated variance of R1b-L23+(i.e. Everything downstream of R1b-L23 and R1b-L23(xL51)) by countries from the haplotype data found in Table-S3 of Myres et al.2010.

Don't know if this is relevant to this discussion but GregRM just posted this link on the Genealogy-DNA mailing list.

https://www.box.com/s/d07b0fc683933615a9c1

http://newsarch.rootsweb.com/th/read/GENEALOGY-DNA/2012-05/1337040189

[alan trowel hands.]

Here is the calculated variance of R1b-L23+(i.e. Everything downstream of R1b-L23 and R1b-L23(xL51)) by countries from the haplotype data found in Table-S3 of Myres et al.2010.



http://i1133.photobucket.com/albums/m582/jeanlohizun/Myresetal2010R1b-L23Variance.jpg

A few observations:

1. There is a clear east to west cline (although the Pakistan result should be ignored).  Very clear indeed. 

2. The results would seem to suggest a route north of the Alps.


3. It is less clear whether eastern/SE Europe or Turkey have priority.  It is perfectly possible from these results that L23 could have occurred in the eastern edges of Europe and simultaneously moved east and west.  The Poland result is very interesting.

JeanL -.  I notice Iberia is missing from the list.  Why is that?

[Maliclavelli]

Of course for me the design is clear. From L150 we have with the SNPs Z2103-Z2105 the Tuscan NA20532 and the Chinese NA18645. Who is wondering of a Chinese close to a Tuscan, go and read my posting on Rootsweb before my banishment when I found a Hui close to me in the STRs and spoke about the paper on the Roman Army at Liqian (the survivors of the defeat at Carrhae who were deported at Merv and after fought against Chinese with the White Huns). Downstream of these we have HG01277 (CLM) and HG01515 (IBS Iberians) with the SNPs Z2108-Z2110.

Downstream L51 we have the Tuscan NA20785 with the SNPs Z2111-Z2119 and downstream we have the Tuscan NA20537 and HG01066 (PUR) with the SNP Z2120.
Then for me it is clear: from Tuscany to Iberia. From Iberia to the Isles and of course to South America.
The Chinese is an outlier, probably a descendant of some Etruscan who was in the Roman Army of Marcus Licinius Crassus.
   

[alan trowel hands.]

What should this demonstrate? The origin of R-L23? The origin of the European subclades, from R-L51 and downstream?
First of all R-L23* are only two, for what I know, i.e. who is also L150-: Romitti (Italy) and Seymour (Great Britain). We have the most part of R-L23 who are L150+. And of course from one of these was born R-L51+. Now, from the ancient map of Argiedude and me, and now from the last map of Richard Rocca, we can see that the centre of R-L51+ is Italy or Italy/France, then that ancestor R-L150+ probably should be found here and not elsewhere, and at this point your variance is worth nothing. We should search the haplotypes closest to R-L51 and subclades.
Your variance is due to many R-L23/L150+ which have nothing to do with our subclades, but which have cumulated variance just for being distant from our European subclades.
Magoon has posted on Rootsweb the new SNPs found in the 1KGP. Many are downstream L150

Z2103 to Z2110

Others are downstream L51

Z2111 to Z2119 and Z2120

Only from L11 descend the subclades P312 and U106 and with this SNP is HG00148.

It seems that Iberia and South America are plenty of R-L11.
Then search on West and not East, you who are born on the Pyrenees.

Nevertheless as a picture of L23 'all' (which is all that its claiming to be) it is a fairly good east to west cline in variance.  

[alan trowel hands.]

There seems to be an older block in the east then a middling block in central Europe then finally a younger group all around the fringes of that central European group.   It looks like a thrust through central Europe from the east and then a dispersal in all directions from that core. 

The origin point would seem most likely to me to be near the western shores of the Black Sea from those results.

[alan trowel hands.]

Here is the calculated variance of R1b-L23+(i.e. Everything downstream of R1b-L23 and R1b-L23(xL51)) by countries from the haplotype data found in Table-S3 of Myres et al.2010.

Don't know if this is relevant to this discussion but GregRM just posted this link on the Genealogy-DNA mailing list.

https://www.box.com/s/d07b0fc683933615a9c1

http://newsarch.rootsweb.com/th/read/GENEALOGY-DNA/2012-05/1337040189

I wonder if we are getting nearer the possibility of SNP counting downstream of M269.

[Maliclavelli]

There seems to be an older block in the east then a middling block in central Europe then finally a younger group all around the fringes of that central European group.   It looks like a thrust through central Europe from the east and then a dispersal in all directions from that core. 

The origin point would seem most likely to me to be near the western shores of the Black Sea from those results.

We don’t know if it is a smooth cline from East to West or from a Centre (Italy/Tuscany) to East and to West in different times. I am waiting that my theories about R1b1* to the R1b1*/R1b1a2* of Mangino (the Tuscan Mancini) are verified (or falsified).

[JeanL]

JeanL -.  I notice Iberia is missing from the list.  Why is that?

I only included haplotypes found in Table-S3, unfortunately Iberian haplotypes did not appear on the Myres et al(2010) study. Also, the Caucasus is as a whole has a pretty big variance, but when individual populations with a sample size of at least 5 are sampled the variance banishes. Almost all of the R1b-L23+ found in the Caucasus is R1b-L23(xL51). This seems to suggest to me, that the R1b-L23(xL51) in the Caucasus might have multiple origins, I'm just not sure if is along language lines(i.e. Turkish vs. Caucasian, etc).

[alan trowel hands.]

Here is the calculated variance of R1b-L23+(i.e. Everything downstream of R1b-L23 and R1b-L23(xL51)) by countries from the haplotype data found in Table-S3 of Myres et al.2010.



http://i1133.photobucket.com/albums/m582/jeanlohizun/Myresetal2010R1b-L23Variance.jpg

Jean- just to check- is that a list of everything that is  downstream of L23?

[JeanL]

Jean- just to check- is that a list of everything that is upstream of L51 and downstream of L23?

Yup, I only excluded R1b-M269(xL23) from the list, so it is everything that is downstream of L23, with L23(xL51) included.

There are a few multi-step mutations found here and there, mostly on populations with big sample sizes like Italy, or Ireland, but they are so few, that they do not affect the overall variance significantly. On the other hand, I can tell you I checked Poland to make sure that their variance was not due to a multi-step mutation being counted as multiple single steps, and they have a very low likelihood of having multi-step mutations.

[Maliclavelli]

There are a few multi-step mutations found here and there, mostly on populations with big sample sizes like Italy, or Ireland, but they are so few, that they do not affect the overall variance significantly.

How are you sure that they are multi-step mutations and not outlier values which demonstrate the mutations for the tangent and not around the modal, then a higher ancientness and variance?

[JeanL]

How are you sure that they are multi-step mutations and not outlier values which demonstrate the mutations for the tangent and not around the modal, then a higher ancientness and variance?

Well first off, whether they are counted as multi-step or as multiple single steps does not significantly change the overall variance of the groups that have them, because the sample sizes are rather large (At least in the populations I have analyzed from the Myres et al.2010 dataset, it is worrisome when it happens in a population of small sample size though). Nonetheless how do I know when it is a multi-step versus a single step, well, if a haplogroup has a mutation than it is 3-4 mutation steps from the modal at any given locus, if it is not a multi-step, one should expect to see at least 1-2 mutations in equally fast loci. Chances are if a haplotype is equal to the modal in almost all loci, but it is 3-4 mutation apart from the modal in one locus, that one is in the presence of a multi-step mutation. There is a small probability that the haplotype is really old, and it just so happens that all other mutations back mutated back to the modal. But Occam’s razor would say otherwise.

[Maliclavelli]

Of course I should see every samples, but your reasoning is likely, even though 3-4 steps are many, specially in middle or slow mutating markers. My haplotype on 12 markers is almost rare (but it finds a pretty close match in a Hui for instance): 12,24,15,10,11-14,12,12,12,13,12 (from 13 recently) 29. My cousin’s husband haplotype finds pretty 100 matches on FTDNA: 12,24,14,11,11-14,12,12,13 (but it could be also 12 or 11),29.
Of course when it is possible I’ll test myself (and my cousin) for these new SNPs. I expect me to be Z2103-Z2105, whereas I expect my cousin to be closer to the cluster which generated L51, and these SNPs could be disjoined.

[alan trowel hands.]

I notice on Dienekes that he is taking out of a study of G that G was more westerly, J southerly and R clades more easterly in the Neolithic.  The lack of R1b in European Neolithic samples to date does seem to hint that it was positions off the beaten track of the early Neolithic movements into Europe from the Levant and Anatolia.  So Dienekes seems to be hanging his hat on 'east' although he hasnt elaborated what he means by that.

[Mikewww]

Here is the calculated variance of R1b-L23+(i.e. Everything downstream of R1b-L23 and R1b-L23(xL51)) by countries from the haplotype data found in Table-S3 of Myres et al.2010.



http://i1133.photobucket.com/albums/m582/jeanlohizun/Myresetal2010R1b-L23Variance.jpg

A few observations:

1. There is a clear east to west cline (although the Pakistan result should be ignored).  Very clear indeed.  

2. The results would seem to suggest a route north of the Alps.


3. It is less clear whether eastern/SE Europe or Turkey have priority.  It is perfectly possible from these results that L23 could have occurred in the eastern edges of Europe and simultaneously moved east and west.  The Poland result is very interesting.

JeanL -.  I notice Iberia is missing from the list.  Why is that?

Thank you, Jean L.  This is much appreciated.

To me it is noteworthy that on pretty decent sample sizes Turkey comes up as .303 while France comes up as .2415.  That's about a 25% difference.  

Poland is also high, on a sample size of 24.  I'm not sure if that is conclusive, but it probably is. Poland seems to be persistent in showing haplogroup and haplotype diversity for R1b.

I've highlighted Alan's comment suggesting a route north of the Alps for R1b.  We don't know, but my opinion is definitely on this side.  I've looked and looked for the Mediterranean routes into Western Europe but I just can't see it in the numbers.  To conjecture a Mediterranean route I have to resort to anecdotal hapltypes.  That could be cherry picking. It's possible, but the weight of the diversity information seems to point to an overland route for the bulk of R1b as we now know it.  Perhaps some explorers came via the Med but the large colonizing settler groups seem to come overland.  I think that matches the common sense of logistics as well.  It's a long way from New England or the Carribean to California but you can make a coastal southern route with a brief crossing in Central America.  The old overland trails like the Oregon Trail carried a lot of people, and then later you have the railroad, etc.  Of course, the riverways are a big part of the hop-skipping through the overland trails.

[razyn]

I've looked and looked for the Mediterranean routes into Western Europe but I just can't see it in the numbers.  To conjecture a Mediterranean route I have to resort to anecdotal haplotypes.  That could be cherry picking. It's possible, but the weight of the diversity information seems to point to an overland route for the bulk of R1b as we now know it.  Perhaps some explorers came via the Med but the large colonizing settler groups seem to come overland.  I think that matches the common sense of logistics as well.  It's a long way from New England or the Carribean to California but you can make a coastal southern route with a brief crossing in Central America.  The old overland trails like the Oregon Trail carried a lot of people, and then later you have the railroad, etc.  Of course, the riverways are a big part of the hop-skipping through the overland trails.

I'm in agreement with your overall argument, but would also point out that Poland is not on the Mediterranean.  If there was a maritime leap to Iberia (whether or not associated with Bell Beaker grave goods) prior to the overland trek of a much more massive breeding population (complete with wives, cattle, Y-DNA haplogroup diversity, etc.), I still don't see why it has to have been Mediterranean.  Present concentrations on Sardinia to the contrary notwithstanding.

I thought this post yesterday by MOESAN was pertinent, if a little confusing:

http://www.eupedia.com/forum/showthread.php?27508-Bell-Beakers-from-Germany-Y-haplogroup-R1b&p=3954

It was almost immediately contradicted by rms2, as you will see if you read it.  But the Dinaric thing was only one of five or six associated traits (physical, linguistic, material culture, Y-DNA).  To me, and I think to MOESAN, the emphasis should be elsewhere in his post -- an unexpected exception to the corded ware rule, that he noticed -- so I'll just highlight that:

Bell Beakers ones was found in one, the Khvalynsk culture in Russia (stage Kurgan III, before -3000?) - if it's true, it's very important, isn't it?

This assertion was pretty clear to me, but his source wasn't -- some unnamed and undated compilation, possibly edited or chaired by Bernard Sergent, and having to do with Indo-European language studies.  Jean M or Alan TH or somebody may know what compilation he's citing, but I don't.  Anyway, Khvalynsk (variously spelled, or transliterated) is a smallish river port way down the Volga, between Samara and Saratov.

[Mikewww]

I've looked and looked for the Mediterranean routes into Western Europe but I just can't see it in the numbers.  To conjecture a Mediterranean route I have to resort to anecdotal haplotypes.  That could be cherry picking. It's possible, but the weight of the diversity information seems to point to an overland route for the bulk of R1b as we now know it.  Perhaps some explorers came via the Med but the large colonizing settler groups seem to come overland.  I think that matches the common sense of logistics as well.  It's a long way from New England or the Carribean to California but you can make a coastal southern route with a brief crossing in Central America.  The old overland trails like the Oregon Trail carried a lot of people, and then later you have the railroad, etc.  Of course, the riverways are a big part of the hop-skipping through the overland trails.

I'm in agreement with your overall argument, but would also point out that Poland is not on the Mediterranean.....

Yes, that's the point I'm trying to make.  I have a hard time seeing how a Mediterranean route fits the genetic diversity of R1b in Europe. In other words, I lean towards the alternative that the majority of R1b in Western Europe came from ancestry that moved overland through Europe.

On the other hand, to say that zero R1b came to Iberia from the Mediterranean is a little hard to swallow, though. These guys seem to move just fine via boat on and along water, both rivers and seas.

[Mikewww]

...
I thought this post yesterday by MOESAN was pertinent, if a little confusing:

http://www.eupedia.com/forum/showthread.php?27508-Bell-Beakers-from-Germany-Y-haplogroup-R1b&p=3954

It was almost immediately contradicted by rms2, as you will see if you read it....

It's a little hard to jump back and forth between simultaneous threads of the same topic, although I think quoting and citing other sources (including forums) is fine.

I see there is a reference to the Khvalynsk culture. Can you re-summarize the pertinent part of the conversation? Moesan lost me.

[razyn]

I can't summarize his argument, but I think I can move this one to the other WF thread, about the R1b Bell Beaker remains in Germany.

Moesan is confusing because he's French, and citing a French publication I can't find, but writing about it in English.  Doesn't mean he doesn't know what he's talking about.  I wouldn't do as well if I wrote this in French.  I still assume the main point is that Bell Beaker ceramics and other burial patterns have been found in the vicinity of Khvalynsk, in a context dated roughly 3,000 BC.  And if so, that's newsworthy.

[JeanL]

Ok so I decided to run yet another analysis of the Myres.et.al.2010 dataset.

Some cautionary comments first:

1)I know 10 STRs is small, but for the purpose of comparison, I think one should look at it.
2)As we will see, the modal of course, gets skewed usually due to population dynamics.

So it goes like this:

Population----Haplogroup-----N----- DYS19—DYS388—DYS389I—DYS389II—DYS390—DYS391—DYS392—DYS393—DYS439—DYS461

Modal(L23+)----L23+----956-----14—12—13—16—24—11—13—13—12—11

Modal(L23xL51)----L23----212-----14—12—13—16—24—11—13—12—12—10

European Modal(L23+)----L23+----812-----14—12—13—16—24—11—13—13—12—11

Western European Modal(L23+)----L23+----632-----14—12—13—16—24—11—13—13—12—11

Western European Modal(L23xL51)----L23----29----14—12—13—16—24—11—13—12—12—10

Caucasus Modal(L23+)----L23+----33----14—12—13—16—24—11—13—12—12—10

Caucasus Modal(L23xL51)----L23----33----14—12—13—16—24—11—13—12—12—10

Turkey Modal(L23+)----L23+----69----14—12—13—16—24—11—13—12—12—10

Turkey Modal(L23xL51)----L23----33----14—12—13—16—24—11—13—12—12—10

So basically the L23xL51 modal is the same for everyone, be them European, or be them West Asian. In turn this modal is 2 mutations different from the modal of all L23+, however as we see, this is mostly due to the vast majority of the sample being European skewing the results. Now, let’s take a look at some haplotypes found in this study:
Individual----Haplogroup---DYS19—DYS388—DYS389I—DYS389II—DYS390—DYS391—DYS392—DYS393—DYS439—DYS461

Turkey #1----L23xL51----15—12—14—15—25—10—12—12—11—10

This haplotype differs from the L23xL51 modal by 7 one-step mutations, and from the L23+ modal by 9 one-step mutations.

Romania----L23xL51----13—12—14—17—25—11—12—12—13—10

This haplotype differs from the L23xL51 modal by 6 one-step mutations, and from the L23+ modal by 8 one-step mutations.

Turkey#2----L23xL51----15—12—12—15—23—10—14—12—12—10

This haplotype differs from the L23xL51 modal by 6 one-step mutations, and from the L23+ modal by 8 one-step mutations.

Turkey#3----L23xL51----14—13—14—16—23—10—13—12—11—10

This haplotype differs from the L23xL51 modal by 5 one-step mutations, and from the L23+ modal by 7 one-step mutations.

Caucasus Kabardinians----L23xL51 ----14—12—13—15—24—10—15—12—11—10

Beware: This haplotype differs from the L23xL51 modal by 3 one-step mutations and 1 double-step mutation, and from the L23+ modal by 5 one-step mutations and one double-step mutation.
Denmark----L23xL51----15—12—14—16—23—11—13—12—10—10

Beware: This haplotype differs from the L23xL51 modal by 3 one-step mutations and 1 double-step mutation, and from the L23+ modal by 5 one-step mutations and one double-step mutation.

Turkey#4----L23xL51----13—12—13—18—25—11—13—12—11—10

Beware: This haplotype differs from the L23xL51 modal by 3 one-step mutations and 1 double-step mutation, and from the L23+ modal by 5 one-step mutations and one double-step mutation.

Turkey#5----L23xL51----15—12—13—16—21—10—13—12—12—10

Beware: This haplotype differs from the L23xL51 modal by 2 one-step mutations and 1 triple-step mutation, and from the L23+ modal by 4 one-step mutations and one triple-step mutation.

So, thus far these really ancient haplotypes all share a signature, they are all descended from the L23xL51 modal, which is evident in the sense that they all share DYS393=12, and DYS461=10.

Now, let’s look at the following haplotypes:

Pakistan----L23xL51----14—12—13—15—25—11—14—12—13—9
This haplotype differs from the L23xL51 modal by 5 one-step mutations, and from the L23+ modal by 5 one-step mutations and one double-step mutation.

Switzerland#1----L23xL51----15—12—13—15—24—10—14—12—13—11

 Switzerland#2----L23xL51----15—12—13—15—24—10—14—12—13—11

These haplotypes differ from the L23xL51 modal by 6 one-step mutations, and from the L23+ modal by 6 one-step mutations.

But here is the interesting part of these haplotypes, they are 7 one-step mutations from the haplotype Turkey#1 which is the most divergent haplotype of all.

[intrestedinhistory]

Here is the calculated variance of R1b-L23+(i.e. Everything downstream of R1b-L23 and R1b-L23(xL51)) by countries from the haplotype data found in Table-S3 of Myres et al.2010.



http://i1133.photobucket.com/albums/m582/jeanlohizun/Myresetal2010R1b-L23Variance.jpg

A few observations:

1. There is a clear east to west cline (although the Pakistan result should be ignored).  Very clear indeed. 

2. The results would seem to suggest a route north of the Alps.


3. It is less clear whether eastern/SE Europe or Turkey have priority.  It is perfectly possible from these results that L23 could have occurred in the eastern edges of Europe and simultaneously moved east and west.  The Poland result is very interesting.

JeanL -.  I notice Iberia is missing from the list.  Why is that?

If it occurred on the eastern fringes of Europe how do R1b rich populations like Assyrians have none of the Northern European component?