World Families Forums

General Forums - Note: You must Be Logged In to post. Anyone can browse. => R1b General => Topic started by: Mike Walsh on January 17, 2011, 07:18:04 PM



Title: R-P312 update variance numbers by country
Post by: Mike Walsh on January 17, 2011, 07:18:04 PM
There are some genetic data points we can review to assess the origin, or at least the original expansion of R-P312. One is variance. I have gathered all of the deep clade tested P312+ haplotypes I can find from volunteer/consumer DNA projects. We have 2295 (actually 2299 as I post this) haplotypes that are 67 markers in length. We have 3400 if you include 12, 25 and 37, etc. haplotypes.

Below is a summary by country with highest variance at the top. Since France is quite interesting, I broke it out by region. I did the same for Italy because I think "Cisalpine Gaul" Italy may be different that Southern Italy. Where there weren't as many haplotypes I aggregated countries. An example is what I call the Nordic area and it includes Norway, Sweden, Finland and Denmark.

Ignore the NVAR (Normalized Variance) column for right now. Please focus on the VAR column. It is the straight Sum of the Variance over the 50 non-multicopy/non-null STR's in FTDNA's first 67. VAR is expressed in terms of percentage points of VAR for all R-P312 (including unknown and New World MDKA's) so it is the Relative Variance as compared to P312 All.

I also display the average and maximum Genetic Distance (GD) for each Country across 67 markers. GD calculations are subject to issues with recLOH's and "infinite allele" markers so I don't recommend using them as the primary tool for analysis but they are another checkpoint and they are fairly easy to understand.

I just displayed the count so you'd get a feel for the quantity of data and its potential representative nature (or not.) There more confirmed deep clade R-P312 than the counts below. For instance, we have 48 haplotypes of 67 length for the Nordic Countries but there are another 22 that haven't upgraded to 67 yet.


COUNTRY/REGION Rel.Var._ N.Rel.Var.__ GD to COUNTRY Modal _ No. Ht's

All____________: Var=1.00, NVar=1.00 @50; AvgGD=14, MaxGD=30 @67 (N=2295)

France ________: Var=1.13, NVar=1.28 @50; AvgGD=15, MaxGD=27 @67 (N=93)
_N.C.France____: Var=1.28, NVar=1.24 @50; AvgGD=17, MaxGD=25 @67 (N=23)
_S.E.France____: Var=1.25, NVar=1.28 @50; AvgGD=15, MaxGD=21 @67 (N=7)
_N.E.France____: Var=1.18, NVar=2.46 @50; AvgGD=15, MaxGD=22 @67 (N=11)
_N.W.France____: Var=1.08, NVar=0.97 @50; AvgGD=15, MaxGD=19 @67 (N=22)
_S.W.France&Pry: Var=0.63, NVar=0.69 @50; AvgGD=12, MaxGD=16 @67 (N=15)

Low Countries__: Var=1.07, NVar=0.88 @50; AvgGD=15, MaxGD=22 @67 (N=25)

East/Cent Euro_: Var=1.05, NVar=1.32 @50; AvgGD=16, MaxGD=30 @67 (N=45)

Spain__________: Var=1.04, NVar=1.17 @50; AvgGD=15, MaxGD=26 @67 (N=58)

England All____: Var=0.99, NVar=1.07 @50; AvgGD=14, MaxGD=28 @67 (N=320)
_W.England_____: Var=1.00, NVar=1.09 @50; AvgGD=14, MaxGD=28 @67 (N=241)
_E.England_____: Var=0.96, NVar=1.02 @50; AvgGD=15, MaxGD=26 @67 (N=79)

Germany _______: Var=0.99, NVar=1.01 @50; AvgGD=15, MaxGD=23 @67 (N=101)

Ireland _______: Var=0.99, NVar=0.87 @50; AvgGD=12, MaxGD=27 @67 (N=577)

Switzerland____: Var=0.94, NVar=1.13 @50; AvgGD=15, MaxGD=21 @67 (N=26)

Alpine Area____: Var=0.93, NVar=1.03 @50; AvgGD=14, MaxGD=22 @67 (N=41)

Wales__________: Var=0.93, NVar=1.17 @50; AvgGD=14, MaxGD=23 @67 (N=64)

Scotland_______: Var=0.93, NVar=0.88 @50; AvgGD=13, MaxGD=27 @67 (N=272)

Nordic Area____: Var=0.90, NVar=1.00 @50; AvgGD=14, MaxGD=24 @67 (N=48)

Italy__________: Var=0.88, NVar=0.87 @50; AvgGD=15, MaxGD=22 @67 (N=29)
_S.Italy&Greece: Var=0.90, NVar=0.97 @50; AvgGD=15, MaxGD=21 @67 (N=13)
_N.Italy_______: Var=0.90, NVar=0.79 @50; AvgGD=14, MaxGD=22 @67 (N=13)


Title: Re: R-P312 update variance numbers by country
Post by: GoldenHind on January 17, 2011, 07:47:49 PM
What, if any, conclusion do you draw from the equal variance in Germany, England and Ireland?


Title: Re: R-P312 update variance numbers by country
Post by: Mike Walsh on January 17, 2011, 08:34:59 PM
What, if any, conclusion do you draw from the equal variance in Germany, England and Ireland?
I don't know what to make of that.  Germany's sample size isn't that large, so perhaps that could change, but I don't really think so.  I looked at the South of Germany versus Middle and don't see much difference.

Rather than look at the murky middle, I try to look at areas that seem significantly different or are unexpected.

I would have thought Scotland would be more like Ireland. Perhaps this indicates that Ireland got hit with one more expansion of R-P312 people than Scotland.

It's most compelling that all of Eastern France and North Central France have high variance but SW France (essentially old Aquitaine) does not.  U152 didn't really make it to SW France, at least in our project MDKA's.  The rest of France is either home or a mix-master for R-P312.

Seems strange that Iberia proper has more variance than the Pyrenees vicinity.  I guess old expectations on the land of the Basques are hard to put away.

The Low Countries have a little more variance than I thought.  Seems to fit on the path from France to England.


Title: Re: R-P312 update variance numbers by country
Post by: alan trowel hands. on January 17, 2011, 08:41:40 PM
If I am understanding this correctly this indicates that in western Europe (is the sample of any value in eastern Europe?) the spread was fast and oldest in the east to north-centre of France and after a short delay spread like lightning to everywhere else adjacent about the same time.  

However, how can we explain an origin for S116 in France? I can only think that it was one of those surfing the front of a wave of L11* and some sort of secondary expansion in France. Myres places that in the middle Neolithic.  If it is later Neolithic in date, it does sort of fit a feeling I have had for a long time that (despite the early Iberian beaker dates) that an origin somewhere like the Rhone or somewhere similar n France would be a far more central diffusion point for both S116 groups and the beaker culture which both (whether or not the two things are connected) penetrated from the far west shores of Iberia to the Danube.  

The question then would be how did L11 or pre-expansion S116 get to France and when?  The results do not favour the southern route via Italy.  However, perhaps S116 really did happen in France and we should really be looking more for its L11* ancestor.  Again, I understand that these have a better showing in central Europe, especially given that the sample tends to be worse there.  

What troubles me is this fits quite well the old beaker model and does not fit at all in any simple way (i.e. without any complex reflux models) with the current 'out of Iberia' (perhaps after an epic jump from the east along the Med.) model that archaeologists have come to largely accept over the last handful of years or so based on review of radiocarbon dates.  

A technical question - if you pool all S116 but the composition of it is very dominated by single clades in a number of areas, will this potentially distort things? France seems odd man out to me in that its S116 is very divided into many clades whereas in other areas one clade predominated i.e. S116* (and other local clades) in Iberia, L21 in Ireland, U152 in Italy and Central Europe etc.  Is there any possibility that this result is being effected by this phenomenon.  

  


Title: Re: R-P312 update variance numbers by country
Post by: Mike Walsh on January 17, 2011, 10:54:55 PM
If I am understanding this correctly this indicates that in western Europe (is the sample of any value in eastern Europe?) the spread was fast and oldest in the east to north-centre of France and after a short delay spread like lightning to everywhere else adjacent about the same time...
The sampling from our DNA projects of Eastern Europe is not very representative so I'm not sure if that is a problem, or if in fact, P312 didn't really expand until he was in Western Europe, or just plain originated there.  I wish had some haplotypes from Anatolia.

I think there is still a fundamental consideration in that R-U106 has high variance in Poland and the Baltic states. ..  so there is a chasm between U106 and P312 out there somewhere where the R-L11 TMRCA for them resided.


Title: Re: R-P312 update variance numbers by country
Post by: Mike Walsh on January 17, 2011, 11:36:36 PM
...A technical question - if you pool all S116 but the composition of it is very dominated by single clades in a number of areas, will this potentially distort things? France seems odd man out to me in that its S116 is very divided into many clades whereas in other areas one clade predominated i.e. S116* (and other local clades) in Iberia, L21 in Ireland, U152 in Italy and Central Europe etc.  Is there any possibility that this result is being effected by this phenomenon.
I don't think that assessing P312 (S116) as a whole is a distortion as ultimately whatever subclade level you are analyzing, there will always be another layer of subclades below it.

However, I've noticed what you are saying also... that STR diversity seems to follow subclade diversity with higher STR variance generally aligning with a "balanced mix" of subclades versus one-sided subclade dominated area. U152 seems to be the oldest subclade of P312 so no place it is not present will have a high variance.. at least it seems.

We do have to peel the onion back one layer at a time. U152 and L21 seem to be fairly well defined, variance-wise.  It's SRY2627 and P312* (et al I should say) that are more perplexing.




Title: Re: R-P312 update variance numbers by country
Post by: Mike Walsh on January 18, 2011, 10:01:42 AM
Vince V drew an interpolated map of P312 coalescence ages according to the Myres study. http://tiny.cc/brmdg

This is a different set of data than our volunteer/consumer DNA projects, but the Myres data has France in a "starring" role, similar to our data (below). SE France is the focus from the Myres data whereas our data would show either SE France or NC France (approximately the same).  Myres did not really get to NC or NW France in their survey.


COUNTRY/REGION Rel.Var._ N.Rel.Var.__ GD to COUNTRY Modal _ No. Ht's

All____________: Var=1.00, NVar=1.00 @50; AvgGD=14, MaxGD=30 @67 (N=2295)

France ________: Var=1.13, NVar=1.28 @50; AvgGD=15, MaxGD=27 @67 (N=93)
_N.C.France____: Var=1.28, NVar=1.24 @50; AvgGD=17, MaxGD=25 @67 (N=23)
_S.E.France____: Var=1.25, NVar=1.28 @50; AvgGD=15, MaxGD=21 @67 (N=7)
_N.E.France____: Var=1.18, NVar=2.46 @50; AvgGD=15, MaxGD=22 @67 (N=11)
_N.W.France____: Var=1.08, NVar=0.97 @50; AvgGD=15, MaxGD=19 @67 (N=22)
_S.W.France&Pry: Var=0.63, NVar=0.69 @50; AvgGD=12, MaxGD=16 @67 (N=15)

Low Countries__: Var=1.07, NVar=0.88 @50; AvgGD=15, MaxGD=22 @67 (N=25)

East/Cent Euro_: Var=1.05, NVar=1.32 @50; AvgGD=16, MaxGD=30 @67 (N=45)

Spain__________: Var=1.04, NVar=1.17 @50; AvgGD=15, MaxGD=26 @67 (N=58)

England All____: Var=0.99, NVar=1.07 @50; AvgGD=14, MaxGD=28 @67 (N=320)
_W.England_____: Var=1.00, NVar=1.09 @50; AvgGD=14, MaxGD=28 @67 (N=241)
_E.England_____: Var=0.96, NVar=1.02 @50; AvgGD=15, MaxGD=26 @67 (N=79)

Germany _______: Var=0.99, NVar=1.01 @50; AvgGD=15, MaxGD=23 @67 (N=101)

Ireland _______: Var=0.99, NVar=0.87 @50; AvgGD=12, MaxGD=27 @67 (N=577)

Switzerland____: Var=0.94, NVar=1.13 @50; AvgGD=15, MaxGD=21 @67 (N=26)

Alpine Area____: Var=0.93, NVar=1.03 @50; AvgGD=14, MaxGD=22 @67 (N=41)

Wales__________: Var=0.93, NVar=1.17 @50; AvgGD=14, MaxGD=23 @67 (N=64)

Scotland_______: Var=0.93, NVar=0.88 @50; AvgGD=13, MaxGD=27 @67 (N=272)

Nordic Area____: Var=0.90, NVar=1.00 @50; AvgGD=14, MaxGD=24 @67 (N=48)

Italy__________: Var=0.88, NVar=0.87 @50; AvgGD=15, MaxGD=22 @67 (N=29)
_S.Italy&Greece: Var=0.90, NVar=0.97 @50; AvgGD=15, MaxGD=21 @67 (N=13)
_N.Italy_______: Var=0.90, NVar=0.79 @50; AvgGD=14, MaxGD=22 @67 (N=13)



Title: Re: R-P312 update variance numbers by country
Post by: Mike Walsh on January 18, 2011, 11:07:30 AM
...A technical question - if you pool all S116 but the composition of it is very dominated by single clades in a number of areas, will this potentially distort things? France seems odd man out to me in that its S116 is very divided into many clades whereas in other areas one clade predominated i.e. S116* (and other local clades) in Iberia, L21 in Ireland, U152 in Italy and Central Europe etc.  Is there any possibility that this result is being effected by this phenomenon.
I don't think that assessing P312 (S116) as a whole is a distortion as ultimately whatever subclade level you are analyzing, there will always be another layer of subclades below it.
.... We do have to peel the onion back one layer at a time. U152 and L21 seem to be fairly well defined, variance-wise.  It's SRY2627 and P312* (et al I should say) that are more perplexing.

Below is a start at peeling the onion back.. an analysis of U152 and it's largest and oldest identified subclade - L2.  You'll see less countries listed but that is not because there are zero haplotypes there, just that the data is getting more and more limiting.  I just don't see the point in doing a statistical calculation on a sample of 2 or 3 when it may be of questionable value under 10 (or more).

COUNTRY/REGION Rel.Var._ N.Rel.Var.__ GD to COUNTRY Modal _ No. Ht's

All U152_______:  Var=1.01, NVar=1.17 @50;  AvgGD=15, MaxGD=30 @67  (N=312)   

France_________:  Var=1.25, NVar=1.68 @50;  AvgGD=16, MaxGD=25 @67  (N=26)   
_NC France_____:  Var=1.51, NVar=1.15 @50;  AvgGD=17, MaxGD=25 @67  (N=7)<<   
_NE France_____:  Var=1.14, NVar=3.06 @50;  AvgGD=14, MaxGD=17 @67  (N=6)<<   

Low Countries__:  Var=1.25, NVar=1.06 @50;  AvgGD=16, MaxGD=21 @67  (N=10)   

Nordic Area____:  Var=1.06, NVar=0.60 @50;  AvgGD=13, MaxGD=20 @67  (N=6)<<   

East/Cent Eur*_:  Var=1.05, NVar=1.38 @50;  AvgGD=17, MaxGD=29 @67  (N=22)   

Germany________:  Var=1.00, NVar=1.23 @50;  AvgGD=15, MaxGD=23 @67  (N=42)   
_Mid Germany___:  Var=1.05, NVar=1.28 @50;  AvgGD=16, MaxGD=18 @67  (N=14)   
_S.Germany_____:  Var=1.01, NVar=1.27 @50;  AvgGD=15, MaxGD=19 @67  (N=17)   

England________:  Var=1.00, NVar=1.15 @50;  AvgGD=15, MaxGD=23 @67  (N=50)

Switzerland____:  Var=0.98, NVar=1.18 @50;  AvgGD=14, MaxGD=22 @67  (N=19)

Alpine Area**__:  Var=0.97, NVar=1.06 @50;  AvgGD=15, MaxGD=22 @67  (N=32)   

Ireland________:  Var=0.97, NVar=0.98 @50;  AvgGD=14, MaxGD=21 @67  (N=16)   

Scotland_______:  Var=0.97, NVar=0.96 @50;  AvgGD=15, MaxGD=23 @67  (N=17)

Italy__________:  Var=0.91, NVar=0.90 @50;  AvgGD=15, MaxGD=22 @67  (N=22)
_N.Italy_______:  Var=0.93, NVar=0.82 @50;  AvgGD=15, MaxGD=22 @67  (N=11)   
_S.Italy_______:  Var=0.85, NVar=0.93 @50;  AvgGD=15, MaxGD=21 @67  (N=8)<<   

Spain & Port.__:  Var=0.88, NVar=0.95 @50;  AvgGD=14, MaxGD=19 @67  (N=7)<<   

* Belarus, Czech Republic (6), Estonia, Hungary (2), Latvia, Lithania, Poland (6), Slovakia, Russia, Ukraine
** Switzerland, Austria & N. Italy
<< small samples

All L2_________:  Var=1.00, NVar=1.09 @50;  AvgGD=16, MaxGD=31 @67  (N=146)

L2 Germany_____:  Var=1.10, NVar=1.41 @50;  AvgGD=15, MaxGD=21 @67  (N=16)   

L2 Italy_______:  Var=1.06, NVar=0.89 @50;  AvgGD=15, MaxGD=22 @67  (N=7)<<

L2 East/Cen Eu_:  Var=1.01, NVar=1.37 @50;  AvgGD=17, MaxGD=27 @67  (N=10)   

L2 France______:  Var=0.96, NVar=0.83 @50;  AvgGD=14, MaxGD=21 @67  (N=16)   
   
L2 England_____:  Var=0.93, NVar=1.20 @50;  AvgGD=15, MaxGD=25 @67  (N=25)


I've expected to see high variance for U152 in Switzerland, Austria and N.Italy (Cisalpine Gaul) or what I've labeled as the "Alpine Area."  I should probably have included SE France and S Germany but you can see those numbers separately for yourself. U152 All doesn't really look "older" in the Alpine Area.  Besides the North of France you can see that Middle Germany (mostly the Rhineland-Palatinate and Rhine Westphalia) has higher variance than South Germany

L2 gives a little different picture than U152 All, but alas, the more you peel the onion back, the less data you have at that level.

Another thing I noticed about the numbers for U152 is East/Central Europe. I've expected high variance there and there is, but not has high as France.  However, loot at the GD's over 67 in East/Central Europe for U152 All. The average GD to the E/C Europe U152 All modal is 17 and the max GD is 29. These are the highest country/region GD's on the board for U152 All. The GD calculation brings 16 multi-copy markers and 425 into play so that seems to have an impact in this case.  I do use modified infinite allele step counting for markers like YCAII, 413 (YCAIII), 464, CDY so that mutes their impact somewhat.

Oh, almost forgot.  Did you notice that U152 has high variance in the Low Countries (Belgium, Luxembourg and the Netherlands)?


Title: Re: R-P312 update variance numbers by country
Post by: Mike Walsh on January 18, 2011, 02:45:36 PM
Below is a start at peeling the onion back.. an analysis of U152...
Here is the next peel of the R-P312 onion .... SRY2627.

So far, L176.2 has been very consistent as being being upstream of both L165 and SYR2627. I am going to show the variance for L165 and the assumed (by me) R-L176.2 All.  Why this is might be important is that L165 is found in Sweden and in a Scots clan that think they have a Scandinavian ancestry.

"VAR" is Relative Sum of the Variance for each respective group as percentage of the Sum of the Variance for R-P312 All (1.0=100%) for 50 non-multicopy/non-null markers.

COUNTRY/REGION Rel.Var._ N.Rel.Var.__ GD to COUNTRY Modal _ No. Ht's

L176.2 All_____:  Var=0.85, NVar=0.93 @50;  AvgGD=14, MaxGD=27 @67  (N=155)   

L176.2/L165____:  Var=0.65, NVar=0.56 @50;  AvgGD=13, MaxGD=22 @67  (N=9)

L176.2/SRY2627_:  Var=0.82, NVar=0.92 @50;  AvgGD=14, MaxGD=29 @67  (N=141)

Spain__________:  Var=0.87, NVar=0.93 @50;  AvgGD=14, MaxGD=20 @67  (N=7) <<   
France_________:  Var=0.86, NVar=1.07 @50;  AvgGD=14, MaxGD=21 @67  (N=14)
Germany________:  Var=0.85, NVar=0.64 @50;  AvgGD=12, MaxGD=20 @67  (N=7) <<   
England________:  Var=0.80, NVar=1.35 @50;  AvgGD=14, MaxGD=21 @67  (N=30)   
Ireland________:  Var=0.73, NVar=0.60 @50;  AvgGD=11, MaxGD=17 @67  (N=15)


The perplexing issue on SRY2627 is that is supposed to be an "Iberian" subclade common among Basques and especially in Catalonia. Our volunteer DNA projects actually have more SRY2627+ guys from England or Ireland than Iberia.

My variance results show Spain, France and Germany all to be of pretty much the same age. Then you add L176.2+ SRY2627- and you see a group that has a lot of Northern European in it.


Title: Re: R-P312 update variance numbers by country
Post by: Mike Walsh on January 19, 2011, 10:47:30 AM
Below is a start at peeling the onion back.. an analysis of U152...
Here is the next peel of the R-P312 onion .... SRY2627.
Here are the R-L21 numbers.

"VAR" is Relative Sum of the Variance for each respective group as percentage of the Sum of the Variance for R-P312 All (1.0=100%) for 50 non-multicopy/non-null markers.

COUNTRY/REGION Rel.Var._ N.Rel.Var.__ GD to COUNTRY Modal _ No. Ht's
All L21________:  Var=0.91, NVar=0.96 @50;  AvgGD=14, MaxGD=28 @67  (N=1174)   

France_________:  Var=1.06, NVar=1.10 @50;  AvgGD=16, MaxGD=21 @67  (N=26)   
_NW France_____:  Var=1.17, NVar=1.25 @50;  AvgGD=16, MaxGD=19 @67  (N=7) <<   
_NC France_____:  Var=1.15, NVar=1.25 @50;  AvgGD=17, MaxGD=21 @67  (N=10)
_NE France_____:  Var=1.07, NVar=1.16 @50;  AvgGD=15, MaxGD=21 @67  (N=4) <<   
_SW Fra &Andorr:  Var=0.70, NVar=0.61 @50;  AvgGD=12, MaxGD=18 @67  (N=5) <<

Wales__________:  Var=0.95, NVar=1.22 @50;  AvgGD=15, MaxGD=22 @67  (N=49)   

England*_______:  Var=0.94, NVar=1.05 @50;  AvgGD=14, MaxGD=27 @67  (N=156)
_West England__:  Var=1.04, NVar=1.02 @50;  AvgGD=15, MaxGD=28 @67  (N=45)   
_East England__:  Var=0.90, NVar=1.03 @50;  AvgGD=14, MaxGD=23 @67  (N=36)

Spain & Port.__:  Var=0.93, NVar=0.87 @50;  AvgGD=14, MaxGD=17 @67  (N=16)   

Ireland**______:  Var=0.90, NVar=0.92 @50;  AvgGD=14, MaxGD=26 @67  (N=332)
_Ulster Ire____:  Var=0.94, NVar=0.94 @50;  AvgGD=14, MaxGD=24 @67  (N=91)   
_Leinster Ire__:  Var=0.89, NVar=1.14 @50;  AvgGD=14, MaxGD=23 @67  (N=44)
_Connacht Ire__:  Var=0.89, NVar=0.89 @50;  AvgGD=14, MaxGD=20 @67  (N=33)   
_Munster Ire___:  Var=0.81, NVar=0.75 @50;  AvgGD=13, MaxGD=27 @67  (N=65)   

Germany________:  Var=0.88, NVar=0.75 @50;  AvgGD=15, MaxGD=21 @67  (N=21)

Scotland***____:  Var=0.84, NVar=0.87 @50;  AvgGD=14, MaxGD=25 @67  (N=188)   
_E.Scotland____:  Var=0.96, NVar=1.08 @50;  AvgGD=15, MaxGD=23 @67  (N=20)
_S.Scotland____:  Var=0.84, NVar=0.81 @50;  AvgGD=13, MaxGD=18 @67  (N=15)   
_Strathclyde Sc:  Var=0.82, NVar=0.78 @50;  AvgGD=14, MaxGD=26 @67  (N=43)   
_N.Scotland____:  Var=0.74, NVar=0.80 @50;  AvgGD=13, MaxGD=22 @67  (N=41)   

* England's administrative regions
** Ireland's old provinces so N.Ireland is in Ulster
*** Strathclyde is old Strathclyde and most of Central. North is the Highland and most of the islands. South is Dumfries, and the Borders and Lothians. East is mostly Grampian and Tay.

I thought I already posted this, but I must have over-written it with my SRY2627 data.

I don't think this is new news, but R-L21 has its highest variance across Northern France but seems younger in SW France.  I only have one long haplotype for SE France.

R-L21 is not absent by any means from Iberia, but it's variance is not high there. England does have greater variance than Ireland and Scotland and the west side of England and Wales have the greatest variance on the British Isles.  There is one strange haplotype in the West Midlands that has DYS481=29. This is +7 of WAMH and +3 of any other R-L21 individual... any feedback on 481? Can it jump multiple steps at once?

Ulster has the greatest variance within Ireland, perhaps not unsurprisingly, but just a few miles across the Irish Sea over in Strathclyde, variance is lower. The East side of Scotland has more variance.  Is this a signal of R-L21 Norse Vikings or is this a signal of more R-L21 incursions from England? or ?


Title: Re: R-P312 update variance numbers by country
Post by: y24 on January 19, 2011, 11:17:49 AM
Below is a start at peeling the onion back.. an analysis of U152...
Here is the next peel of the R-P312 onion .... SRY2627.
The East side of Scotland has more variance.  Is this a signal of R-L21 Norse Vikings or is this a signal of more R-L21 incursions from England? or ?


What is the variance of Norwegian and Scandinavian L21 compared to east coast Scotland L21, I don't see Scandinavian L21 on the chart? Does more variance signify an older population?


Title: Re: R-P312 update variance numbers by country
Post by: MHammers on January 19, 2011, 12:57:55 PM
Mike,

Nice breakdown of the data.  It looks like the L11 ancestor to P312 possibly came right up the Danube and then across the Rhine into France, before branching out. 

Is there any chance those haplotypes could be run through Ken's Generations5 spreadsheet to see how many generations/years we're looking at?


Title: Re: R-P312 update variance numbers by country
Post by: Mike Walsh on January 19, 2011, 01:37:32 PM
.... Is there any chance those haplotypes could be run through Ken's Generations5 spreadsheet to see how many generations/years we're looking at?
Yes, the whole haplotype file is available at the R-P312 Yahoo Group files section.  I can send a slimmed down version of it to whoever will do an analysis.  Anatole K now has it, for example.
 
I'm not well versed on Ken's program.  I know that Vince V and Tim J have a modified version that may be easier to use. Do you want the data?  67 length haplotypes only?  Just tell me what you want and where to email it via a private message.

There is a guy on the R-L21 Yahoo Group, Alex W, who does a great job of running the R-L21 ht's through a diagramming tool and making speculative deep ancestral and cluster relationships easy to understand.  I'm not sure his "clusters" are better than anyone else's but people love the diagrams and it gives them reasons to test out to 67 markers and do downstream SNP testing.


Title: Re: R-P312 update variance numbers by country
Post by: alan trowel hands. on January 19, 2011, 01:59:43 PM
France really keeps topping the charts.  

The data for variance overall does not at all support the idea of the spread of S116 along the Med.


Title: Re: R-P312 update variance numbers by country
Post by: Mike Walsh on January 19, 2011, 02:59:51 PM
... The data for variance overall does not at all support the idea of the spread of S116 along the Med.
I keep expecting to find some Mediterranean path to Iberia with much higher variance there or more R-L23* there in Iberia or something.  I'm just not finding it.

I look at this Bell Beaker migration map and find it hard to make a match with genetics.
(http://secher.bernard.free.fr/DNA/Diffusion_campaniforme.jpg)

The only thing that makes sense to me right now is a picture of the North America. European (IE cultural descendants, BTW) maritime explorers formed colonies on the coasts, naturally, the eastern coast. They were looking for trade routes, precious metals, etc.  Some, ironically Spanish speaking, hit the lower lattitudes, crossed the narrowest part of the continent (Central America) and moved up the western side of the continent towards California. Explorers criss-crossed the continent and small colonies were started, but that's about it. The vast majority of the continent's land mass and resources remained indigenous.

When the colonies built enough infrastructure and sophistication, the door was open for larger inflows of Europeans who eventually expanded in-mass, overland, east to west..  cattle herders, farmers, with their horses, wagons and the whole bit. My guess is shipping by sea could not provide enough transportation for a massive migration so the colonies had to "build out" first, as staging points for the overland mass migration.

These are just out-loud thoughts.


Title: Re: R-P312 update variance numbers by country
Post by: alan trowel hands. on January 19, 2011, 04:47:13 PM
... The data for variance overall does not at all support the idea of the spread of S116 along the Med.
I keep expecting to find some Mediterranean path to Iberia with much higher variance there or more R-L23* there in Iberia or something.  I'm just not finding it.

I look at this Bell Beaker migration map and find it hard to make a match with genetics.
(http://secher.bernard.free.fr/DNA/Diffusion_campaniforme.jpg)

The only thing that makes sense to me right now is a picture of the North America. European (IE cultural descendants, BTW) maritime explorers formed colonies on the coasts, naturally, the eastern coast. They were looking for trade routes, precious metals, etc.  Some, ironically Spanish speaking, hit the lower lattitudes, crossed the narrowest part of the continent (Central America) and moved up the western side of the continent towards California. Explorers criss-crossed the continent and small colonies were started, but that's about it. The vast majority of the continent's land mass and resources remained indigenous.

When the colonies built enough infrastructure and sophistication, the door was open for larger inflows of Europeans who eventually expanded in-mass, overland, east to west..  cattle herders, farmers, with their horses, wagons and the whole bit. My guess is shipping by sea could not provide enough transportation for a massive migration so the colonies had to "build out" first, as staging points for the overland mass migration.

These are just out-loud thoughts.

There is no doubt that the story of European R1b1b2 does not match the  current thinking on the spread of beakers.  Either the archaeologists or the geneticists have got their models wrong OR there is no correlation.  However, it is only the preferred beaker model of the last handful of years that gives this contradiction.  The preferred models for beaker culture before the out of Iberia model gained popularity were actually much more compatible with the DNA evidence.  I have seen models come and go and come again in again in almost circular fashion so with archaeology it is naive to think the latest most popular model is the correct one.  I would not be at all surprised if the out of Iberia beaker model is overturned by the future discovery of older dates in France or further east into central Europe.  


Title: Re: R-P312 update variance numbers by country
Post by: IALEM on January 19, 2011, 04:59:13 PM
height=85]http://secher.bernard.free.fr/DNA/Diffusion_campaniforme.jpg[/img]



When the colonies built enough infrastructure and sophistication, the door was open for larger inflows of Europeans who eventually expanded in-mass, overland, east to west..  cattle herders, farmers, with their horses, wagons and the whole bit. My guess is shipping by sea could not provide enough transportation for a massive migration so the colonies had to "build out" first, as staging points for the overland mass migration.

These are just out-loud thoughts.
However the key factor for larger inflows of Europeans were not infrastructure in America but technological changes in naval construction taht allowed transatlantic voyages with ships that were larger and faster. Maybe we should look for a similar technological change rather than time for building colonies


Title: Re: R-P312 update variance numbers by country
Post by: alan trowel hands. on January 19, 2011, 05:49:31 PM
There is no question that R1b1b2 was spread by some sort of east to west movement, and the data relating to L11 derived clades would seen to suggest that the route was though central Europe.  On that journey it seems that France was the point of expansion for U152 and L21 and possibly other clades.  So, it seems we are looking for an east-west movement with a period of especially successful expansion occurring in and around the eastern half of France.  

This does at least vaguely match the Myres model of some L11* arriving in the LBK, later doing not so great/declining before then suddenly having a secondary expansion in the middle Neolithic of France and immediately adjacent areas.  These middle Neolithic cultures may have expanded from in and around France into Spain, Italy etc, thereby explaining away how L11 and even S116 cuts across the older earlier Neolithic LBK-Cardial divide.  Perhaps the rough times at the end of the LBK means that much of the early story of L11's spread is lost and what we see is the results of its re-expansion in the middle Neolithic.  

In terms of dating, the middle Neolithic cultures of c. 4000BC are not radically older than the earliest beaker dates which reach back to around 3000BC in Iberia.  So, I dont think the middle Neolithic model can be rejected simply on the basis of dating.

It certainly also has the advantage of getting us away from the mental gymnastics needed to explain an east-west genetic move with a west-east archaeological culture (a culture which few archaeologists over the last 40 years have seen as being of demographic importance).  There is a lot in the Myres middle Neolithic expansion model that could fit.  The middle Neolithic is interesting because not only did it seem to herald the revival of settlement in area where it had long been settled by LBK farmers but it also saw groups cross the LBK-Cardial divide.  Finally it also saw the expansion of farming into areas that had not been settled by farmers in the earlier Neolithic thrusts of LBK and Caridal cultures, including all the Atlantic shores of mainland Europe from Galicia to Scandinavia, much Spain away from the small Cardial coastal strip, a chunk of the interior of France, the Alpine areas, as well as the isles.  There was a lot of western Europe that was not settled in the first thrust of the Neolithic and were probably settled by their descendants in the middle Neolithic, perhaps after they had arrested a period of serious decline and undergone a new expansion marked by descendant middle Neolithic cultures.  

So, its a very important period in the settlement history of a huge chunk of Europe. I was certainly led to believe that many of the middle Neolithic cultures displayed clear traits of distant LBK roots.  So, again perhaps there was a spread that included L11* and in hard times in the late LBK much of it died out, only for the remnants to reemerge in successor cultures in the middle Neolithic in the form of U152, L21 etc.  

I just wish there was a massive up to date book on the subject in English.  I suspect much of it is in French and German and to a lesser extent Spanish and Italian journals.  


Title: Re: R-P312 update variance numbers by country
Post by: Mike Walsh on January 19, 2011, 06:23:37 PM
When the colonies built enough infrastructure and sophistication, the door was open for larger inflows of Europeans who eventually expanded in-mass, overland, east to west..
However the key factor for larger inflows of Europeans were not infrastructure in America but technological changes in naval construction taht allowed transatlantic voyages with ships that were larger and faster. Maybe we should look for a similar technological change rather than time for building colonies
I acknowledge that technological advances were a part of it, but the larger ships needed docks, paying passengers and freight, etc. to be viable. There needed to farms, plantations and people with a enough minimum population mass to make it worthwhile.  I don't disagree with you, but I think the colonies had to be poised to handle the commerce and the growth to come.

... but I think this brings out an interesting point.  Which was a bigger factor in European population explosion in North America?  1) Was it a mass of immigrants or 2) the higher than normal (compared to Europe) growth in families thereafter? There was a reason people were coming to America. Economic and religious distress, famines, wars, etc. back home.  On the other hand, there was plenty of land in America and the wars were mostly just battles with indigenous peoples.  I suspect the higher growth rate in the Americas had more influence on the total population than the actual inflow of people. Has anyone looked at this?

This could get back to Alan's point and others' about R-L11* showing up, perhaps with the LBK, but as it petered out, a secondary expansion out of Western or Central Europe triggered the population explosion that spread R-P312 and R-U106.


Title: Re: R-P312 update variance numbers by country
Post by: alan trowel hands. on January 19, 2011, 06:25:54 PM
following on from that, it strikes me that L11* is crucial to a Myres type model.  Is there enough L11* to actually reliably calculate is variance?  It strikes me that if L11* is scarce we might not be sampling its full variety and that could mean that we are underestimating its age.  I say this because clearly a Myres type model of L11* arriving in the LBK but the downstream forms only first expanding in the Neolithic L11* to be perhaps over 1000 years older than U152 etc.  That would mean it should be about an eighth older than any of its subcludes.  


Title: Re: R-P312 update variance numbers by country
Post by: Mike Walsh on January 19, 2011, 07:00:24 PM
Below is a start at peeling the onion back.. an analysis of U152...
Here is the next peel of the R-P312 onion .... SRY2627.
Here are the R-L21 numbers.
There is one more large piece of the onion to peel - R-P312*.  Or rather it is not one peel but all of the other left over peels.. we just don't know their subclades. That being said I've questioned the validity in doing calculations for paragroups in the past and I still do.  However, in case this is helpful, here it is.

COUNTRY/REGION Rel.Var._ N.Rel.Var.__ GD to COUNTRY Modal _ No. Ht's

All P312*______:  Var=1.01, NVar=1.13 @50;  AvgGD=15, MaxGD=27 @67  (N=337)

Ireland________:  Var=1.17, NVar=1.26 @50;  AvgGD=14, MaxGD=26 @67  (N=31)

Spain & Port.__:  Var=1.13, NVar=1.62 @50;  AvgGD=16, MaxGD=26 @67  (N=38)

France_________:  Var=1.11, NVar=1.03 @50;  AvgGD=15, MaxGD=28 @67  (N=25)

East/Cen Eur___:  Var=1.09, NVar=1.19 @50;  AvgGD=16, MaxGD=23 @67  (N=15)

Scotland_______:  Var=0.99, NVar=1.31 @50;  AvgGD=14, MaxGD=21 @67  (N=17)

Nordic Area____:  Var=0.95, NVar=1.23 @50;  AvgGD=15, MaxGD=24 @67  (N=16)

England________:  Var=0.94, NVar=0.84 @50;  AvgGD=14, MaxGD=25 @67  (N=65)

Germany________:  Var=0.91, NVar=0.91 @50;  AvgGD=15, MaxGD=24 @67  (N=27)

Low Countries__:  Var=0.81, NVar=0.63 @50;  AvgGD=13, MaxGD=18 @67  (N=9) <<

These samples are really small but here they are:

Wales__________:  Var=0.80, NVar=1.20 @50;  AvgGD=14, MaxGD=17 @67  (N=5)   

Italy__________:  Var=0.64, NVar=0.53 @50;  AvgGD=12, MaxGD=14 @67  (N=5) <<

Alpine Area____:  Var=0.59, NVar=0.68 @50;  AvgGD=12, MaxGD=14 @67  (N=5) <<

<< smaller samples


This picture does look different as France is not on top as it is with the identified subclades. Ireland and Spain are on top, which should make the Milesians happy. http://en.wikipedia.org/wiki/Milesians_%28Irish%29 (http://en.wikipedia.org/wiki/Milesians_%28Irish%29)

Could the results above point to two (or three more) large unidentified subclades moving into Ireland? One coming up the Atlantic with the trade zone, a second coming across from the continent and France and perhaps a third coming from the east as Scandinavians.  In this sense, Ireland would gain the highest variance of R-P312* because it has the best balance of unidentified subclades.

I am concerned about the sample sizes.  I did the same calculations but with 37 (and above) length haplotypes so I picked up larger samples below. There are only 25 non-multicopy markers in the first 37.

Spain & Port.__:  Var=1.31, NVar=1.59 @25  (N=61)   
East/Cen Eur___:  Var=1.16, NVar=1.51 @25  (N=23)
France_________:  Var=1.16, NVar=1.38 @25  (N=43)
England________:  Var=1.12, NVar=1.06 @25  (N=80)
Ireland________:  Var=1.07, NVar=1.05 @25  (N=38)
Alpine Area____:  Var=1.02, NVar=1.07 @25  (N=7) <<   
Germany________:  Var=1.00, NVar=1.06 @25  (N=43)
Scotland_______:  Var=0.96, NVar=1.07 @25  (N=20)
Nordic Area____:  Var=0.89, NVar=0.98 @25  (N=17)

Spain/Portugal stays on top and but Ireland drops.


Title: Re: R-P312 update variance numbers by country
Post by: Mike Walsh on January 19, 2011, 07:27:49 PM
Below is a start at peeling the onion back.. an analysis of U152...
Here is the next peel of the R-P312 onion .... SRY2627.
Here are the R-L21 numbers.
There is one more large piece of the onion to peel - R-P312*.
One more peel.... I broke P312* into sections - North-South Cluster and non-North-South Cluster.

COUNTRY/REGION Rel.Var._ N.Rel.Var.__ GD to COUNTRY Modal _ No. Ht's
P312* N-S All__:  Var=0.92, NVar=0.85 @50;  AvgGD=13, MaxGD=24 @67  (N=49)

P312* Non N-S__:  Var=1.00, NVar=1.14 @50;  AvgGD=15, MaxGD=27 @67  (N=288)

The N-S cluster appears to be a big chunk of R-P312* and it is spread out very evenly across England, France and Spain and then scattered all over including Scandinavia.  My guess it is really common in Spain and France.


Title: Re: R-P312 update variance numbers by country
Post by: GoldenHind on January 19, 2011, 09:12:43 PM
Quote
Below is a start There is one more large piece of the onion to peel - R-P312*.  Or rather it is not one peel but all of the other left over peels.. we just don't know their subclades. That being said I've questioned the validity in doing calculations for paragroups in the past and I still do.  However, in case this is helpful, here it is.

COUNTRY/REGION Rel.Var._ N.Rel.Var.__ GD to COUNTRY Modal _ No. Ht's

All P312*______:  Var=1.01, NVar=1.13 @50;  AvgGD=15, MaxGD=27 @67  (N=337)

Ireland________:  Var=1.17, NVar=1.26 @50;  AvgGD=14, MaxGD=26 @67  (N=31)

Spain & Port.__:  Var=1.13, NVar=1.62 @50;  AvgGD=16, MaxGD=26 @67  (N=38)

France_________:  Var=1.11, NVar=1.03 @50;  AvgGD=15, MaxGD=28 @67  (N=25)

East/Cen Eur___:  Var=1.09, NVar=1.19 @50;  AvgGD=16, MaxGD=23 @67  (N=15)

Scotland_______:  Var=0.99, NVar=1.31 @50;  AvgGD=14, MaxGD=21 @67  (N=17)

Nordic Area____:  Var=0.95, NVar=1.23 @50;  AvgGD=15, MaxGD=24 @67  (N=16)

England________:  Var=0.94, NVar=0.84 @50;  AvgGD=14, MaxGD=25 @67  (N=65)

Germany________:  Var=0.91, NVar=0.91 @50;  AvgGD=15, MaxGD=24 @67  (N=27)

Low Countries__:  Var=0.81, NVar=0.63 @50;  AvgGD=13, MaxGD=18 @67  (N=9) <<

These samples are really small but here they are:

Wales__________:  Var=0.80, NVar=1.20 @50;  AvgGD=14, MaxGD=17 @67  (N=5)   

Italy__________:  Var=0.64, NVar=0.53 @50;  AvgGD=12, MaxGD=14 @67  (N=5) <<

Alpine Area____:  Var=0.59, NVar=0.68 @50;  AvgGD=12, MaxGD=14 @67  (N=5) <<

<< smaller samples


This picture does look different as France is not on top as it is with the identified subclades. Ireland and Spain are on top, which should make the Milesians happy. http://en.wikipedia.org/wiki/Milesians_%28Irish%29 (http://en.wikipedia.org/wiki/Milesians_%28Irish%29)

Could the results above point to two (or three more) large unidentified subclades moving into Ireland? One coming up the Atlantic with the trade zone, a second coming across from the continent and France and perhaps a third coming from the east as Scandinavians.  In this sense, Ireland would gain the highest variance of R-P312* because it has the best balance of unidentified subclades.

I am concerned about the sample sizes.  I did the same calculations but with 37 (and above) length haplotypes so I picked up larger samples below. There are only 25 non-multicopy markers in the first 37.

Spain & Port.__:  Var=1.31, NVar=1.59 @25  (N=61)   
East/Cen Eur___:  Var=1.16, NVar=1.51 @25  (N=23)
France_________:  Var=1.16, NVar=1.38 @25  (N=43)
England________:  Var=1.12, NVar=1.06 @25  (N=80)
Ireland________:  Var=1.07, NVar=1.05 @25  (N=38)
Alpine Area____:  Var=1.02, NVar=1.07 @25  (N=7) <<   
Germany________:  Var=1.00, NVar=1.06 @25  (N=43)
Scotland_______:  Var=0.96, NVar=1.07 @25  (N=20)
Nordic Area____:  Var=0.89, NVar=0.98 @25  (N=17)

Spain/Portugal stays on top and but Ireland drops.

As you acknowledge, P312* is a paragroup not a subclade, which people sometimes forget. As I understand it, it is highly unlikely that the ancestors of those currently identified as P312* have had no SNPs over the course of several thousand years. This simply means that most if not all P312* individuals have differing subclades that not yet been identified. These subclades may well be as different in age and distribution as that of (for example) L21 from U152, or L21 from SRY2627. We have no idea of what the numbers are for these various unidentified subclades. Therefore my opinion is that average variance of those P312* individuals currently in the project becomes essentially meaningless, and potentially misleading, to the extent that it leads people to look at them as if they were just another subclade.  
I know you pointed this out in your introduction, but I thought a further explanation might be helpful to some readers.


Title: Re: R-P312 update variance numbers by country
Post by: Maliclavelli on January 20, 2011, 05:08:19 AM
There is a great toil, above all from people from Central and North Europe, to demonstrate that, if they are R-U152,  their origin cannot be from Italy, even though Italy has the high percentage of this haplogroup and, above all, an uniform distribution of it: overall half of R.  Even though my theory was (and is) of an Italian Refugium of R1b1* to R-L51, and I have always said that its subclades was born in Central Europe after the migration of R-L51 from Italy, I think I have something to say also on this.
Once the presence of all the pathway of R in Italy, more than in every other country, was justified by saying that Italy had received a migration from all over the world and that this didn’t demonstrate anything. Now the apparent more variance in other countries than Italy does mean that those countries are at the origin of the haplogroup. A clear example of incoherence, perhaps bad faith, probably of racism.
If Italy had been what these said Celts are saying, Italy should have had the highest variance, but in this case this would have meant that Italy was that land of immigration that everybody has always thought and said.
The highest variance of France of R-U152* is due just to the fact that France has received immigrations from all over the Europe (and now from all over the world) being at the cross of all the ways in Europe and the lowest variance in Italy demonstrates just that here lives who descends from the first inhabitants. This is demonstrated also by the total absence of a recent haplogroup of R born in North Western Europe: R-L21.
The recent discovery by Vizachero of an R1b1b2* (Mangino) with the highest variance in this haplogroup and linked closely to the R1b1/V88-/L389+, and probably with some SNP that breaks the chain of SNPs around M269, demonstrates that every theory, above all about those countries like Italy less tested than others of Western Europe, is premature. Anyway only the ancient DNA will solve our questions and I am confident like never before.


Title: Re: R-P312 update variance numbers by country
Post by: Mike Walsh on January 20, 2011, 05:46:00 PM
...The highest variance of France of R-U152* is due just to the fact that France has received immigrations from all over the Europe (and now from all over the world) being at the cross of all the ways in Europe and the lowest variance in Italy demonstrates just that here lives who descends from the first inhabitants. ....
I've been looking for R-U152 haplotypes from Italy, Switzerland and Austria (well, France and Germany too) to see if the variance in N.Italy is high like I use to always think it was.  I'm not find any variance numbers in studies on this, just frequency data. Does anybody have this? I was told the Underhill study found R-U152 was "ancient" in Italy, but I hope they were evaluating diversity and not just frequency.

Tibor F calculated the variance of R-U152 from the Myres published haplotypes.  Here are the top countries.
(Top countries for R-U152 variance)

Germany_____ 2.42 (N=33)
England_____ 2.33 (N=9)
Slovenia____ 2.33 (N=6)
Turkey______ 2.33 (N=3)
Poland______ 2.20 (N=3)
Slovakia____ 2.14 (N=5)
France______ 2.11 (N=19)
Italy_______ 2.00 (N=57)
Greece______ 1.86 (N=7)
Switzerland_ 1.79 (N=29)


Most of the countries have a limited number of ht's, but Italy, Switzerland, Germany and perhaps France have enough to consider.

I'm just not detecting high variance in Italy for R-U152.  This doesn't mean that R-L23* or R-L51* was not there long before U152, but it's beginning to look like the majority of R-M269 in Italy came from the north.

Even if France is a cross-roads or destination for R-U152 types, it shouldn't have more variance than the original source location so between north** of the Alps or south of the Alps, the south seems to be eliminated as the source for U152.

** EDIT: What I  meant is just outside the Alps and Cisalpine Gaul.


Title: Re: R-P312 update variance numbers by country
Post by: alan trowel hands. on January 20, 2011, 08:15:14 PM
One thing I find interesting is that some calculations find S116* to be younger than U152.  That (unless its down to confidence intervals etc) suggests the possibility that the lines that are negative for other SNPs have a common ancestor that is significantly more recent than the S116 SNP, at least by quite a few centuries.  This implies that the early shared ancestral lines of these L21/U152 etc negative S116 group of lineages did not prosper at first.  There seems to be a period of less success in the window between S116 and I152 SNPs.  Possibly it is all a case of the seed falling on fertile ground.  The U152 and L21 lines simply stumbled upon opportunity.  I am not sure about U152 but in a mid Neolithic scenario L21 would have found some virgin territory settled only by hunters around coastal northern France.. In a similar period it is possible that U152 could have benefited from the poorly Neolithicised area around the Alps and in central France where neither the LBK or the Cardial waves had reached with any strength. It is also true that much of Iberia away from the south and west coasts was remote from the main Cardial waves there.  Finally, most of the coastal areas of what is now Germanic Europe was also not Neolithicised until the middle Neolithic.  

Maybe the clade explosion we see among L11 is down to the sudden secondary expansion of Neolithic peoples. Some think France (which was only partly Neolthicised by LBK and Cardial cultures) had a major period of secondary expansion in the middle Neolithic that was responsible for the first true farming cultures in north coastal and part of interior France.  I think I recall it being suggested that these French middle Neolithic groups expansion could have also impacted into interior and northern Iberia and parts of Italy.  Is it possible that this is the link with the peak variance for some clades falling into France?  I certainly struggle to find other periods where France stands out as some sort of centre of expansion.  

The middle Neolithic seems to have been a period where good climate and clever adaptation of farming practices allowed expansion away from the limited areas the rather choosey LBK (they like Loess soils or failing that sands and gravel subsoils) and Cardial settlers had settled (and failed to expand beyond for many centuries). The middle Neolithic was the 2nd post-hunter-gatherer demographic window after the 'spread of farming 1000 years or so earlier. So if you look to match demographic explosions in DNA with windows of opportunity then the middle Neolithic is certainly a major one.  It has the advantages of being 1-2000 years later than the spread of farming which I suppose closes some of the disparity with the variance dates suggested.  

The geography of the places where these opportunities were highest in terms of territory that was essentially virgin with regards to fully developed farming cultures is interesting too - the isles, north and NW coastal fringe France, much of interior France, the low Countries, the Alpine area of Europe (including parts of north Italy), Iberia away from the south or west coasts, Denmark etc.  These area all areas where L11 clades are elevated.  In terms of the map of Europe and even looking at subdivisions of it, it seems R1b1b2 (mainly L11 clades) increases where a strong farming economy was only established in the middle Neolithic.

So, did L11 suddenly explode with the middle Neolithic window of opportunity and create the areas of elevated R1b1b2 we see now?  If so, what were it roots?  There is no strong evidence that the middle Neolithic groups were outsiders.  I think in general these groups are seen as descendants of LBK groups who had adapted to more marginal conditions and perhaps absorbed some local hunters.  




Title: Re: R-P312 update variance numbers by country
Post by: alan trowel hands. on January 20, 2011, 08:33:44 PM
...and I forgot to say.  The correlation between elevated R1b1b2 (well L11 anyway) and areas that were not settled by real farming cultures until the middle Neolithic was misinterpreted as down to hunter gatherer survival a few years ago before the phylology of R1b1b2 was better understood.  You can see how that would come about.  The map of R1b1b2 frequency is almost the reverse of the map of early Neolithic cultures.  R1b1b2 tends to be strongest where the early Neolithic didnt spread to.  So people put 2 and 2 together and got R1b=hunters.  However what I am now saying is that while the observation that R1b1b2 strength is almost inversely correlated with early Neolithic settlement is essentially true, the real reason is not hunter-gatherer survival but middle Neolithic settlement. 

..and the reason why L11 clades are poorly correlated with the LBK spread?  They didnt exist at the time.  Obviously to expand in the middle Neolithic L11 or its ancestors had to have been on the scene and to have been in a particularly advantageous spot.


Title: Re: R-P312 update variance numbers by country
Post by: MHammers on January 20, 2011, 08:47:38 PM
Here are some calculations I ran using Ken N.'s Generations5 interclade calculator.  The data comes from the Yahoo P312 project.  Thanks to Mikewww, Ken, and Vince V. for making these resources available.  These are all 67 marker ht's with only 50 used in the spreadsheet.

All P312 (Generations/Sigma generations , Gx30=years ago , n=)
France - 149/11, 4440, 93
Iberia - 129/9.1, 3870, 66
Germany - 126/11, 3780, 103
Scandinavia - 113/10, 3390, 48
Alpine(Switzerland, North Italy, Austria) - 111/8.7, 3330, 42)

Here are some interclade estimates using W. Europe P312*, n=105, as the ancestor clade or sample A on the spreadsheet.  The following were entered seperately as sample B.  I'm still trying to grasp the interclade concept, so hopefully this is the right approach.

Iberia L176 - 132/42, 3960, 8
France L21 - 136/43, 4080, 27
Iberia L21 - 126/41, 3780, 16
France U152 - 163/50, 4890, 26
Alpine U152 - 133/43, 3990, 32

I plan on running some more later.



Title: Re: R-P312 update variance numbers by country
Post by: MHammers on January 21, 2011, 08:32:28 PM
Some more TMRCA's using 30 yrs. per generation.  I intend to look at the British Isles, once I can modify my copy of the spreadsheet to accommodate the much larger samples.

All P312
Eastern Europe G-127/10, 3810, n-28

All L21
France L21 136/20, 4080, n-27
Iberia 134/17, 4020,  n-16
Germany 127/14, 3810, n-26
E Eur 104/16, 3120, n-16
Scandinavia 101/12, 3030, n-22

All U152
Low Countries 157/15, 4710, n-10
France 151/9.5, 4530, n-26
Iberia 135/21, 4050, n-6
Scandinavia 135/16, 4050, n-6
South Italy 133/20, 3990, n-8
E Eur 127/10, 3810, n-30
Germany 125/10, 3750 n-42
Alpine 116/10, 3480, n-32

All L176
Iberia 115/14, 3450, n-8
Germany 112/15, 3360, n-7
France 112/13, 3360, n-14

Interclade L51*/L11* (Ht 35 project)
Mostly of Central and Western Europe descent 167/51, 5010, n-12/35

The oldest so far.  If you add  or subtract the sigma you get between 3480-6650 years which spans from the middle neolithic to the bronze age.

Interclade L11*/Western Europe P312(includes P312, P312*, and P312** from Yahoo project)
136/44, 4080, n-35/105


Title: Re: R-P312 update variance numbers by country
Post by: GoldenHind on January 22, 2011, 12:01:00 AM
Here are some calculations I ran using Ken N.'s Generations5 interclade calculator.  The data comes from the Yahoo P312 project.  Thanks to Mikewww, Ken, and Vince V. for making these resources available.  These are all 67 marker ht's with only 50 used in the spreadsheet.

All P312 (Generations/Sigma generations , Gx30=years ago , n=)
France - 149/11, 4440, 93
Iberia - 129/9.1, 3870, 66
Germany - 126/11, 3780, 103
Scandinavia - 113/10, 3390, 48
Alpine(Switzerland, North Italy, Austria) - 111/8.7, 3330, 42)

Here are some interclade estimates using W. Europe P312*, n=105, as the ancestor clade or sample A on the spreadsheet.  The following were entered seperately as sample B.  I'm still trying to grasp the interclade concept, so hopefully this is the right approach.

Iberia L176 - 132/42, 3960, 8
France L21 - 136/43, 4080, 27
Iberia L21 - 126/41, 3780, 16
France U152 - 163/50, 4890, 26
Alpine U152 - 133/43, 3990, 32

I plan on running some more later.



Sorry, I do not understand these figures. Could you knidly explain what the three sets of numbers represent, and what they signify?


Title: Re: R-P312 update variance numbers by country
Post by: MHammers on January 22, 2011, 11:17:55 AM
@GoldenHind
The larger number is the generations to the most recent common ancestor for that sample.  The second smaller number is the sigma or standard deviation in generations (for example + or - 500 years).  The third number is actual years, generations x 30 yrs. per generation.  N= sample size.

I also ran a revised interclade for U152 which I think the gives a much better estimation of its age.  The previous interclade for France only had a sample of 26 and contained undifferentiated and downstream (U152 only) types.  The P312 sample had 6 undifferentiated, but no known downstream types.  This time, I used 99 P312* and P312** with 93 U152* and U152**.  This is all of the Western Europe sample which has the highest variance and diversity, so it is a better candidate than say Eastern Europe for the ancestral population.  By using the * and ** members it should be a better representation for the ancestral populations without the undifferentiated members who may actually be L21, L176, L2, etc.
 
I got G=137/44 or 4110 years which is much more in line with the other estimates.  With France having the highest variance and intraclade for U152, it is likely the origin is there or near the Low countries. A better explanation is, if I try to find the age of L21 from somewhere and include M222, a younger subclade with a distinct haplotype from most other L21, it will put the age of L21 higher than it actually is in an interclade estimate.


Title: Re: R-P312 update variance numbers by country
Post by: alan trowel hands. on January 22, 2011, 03:11:02 PM
The L21 date is about 500 years too young if a beaker model is used.  The latest L21 could have emerged from S116 if is associated with the beaker dispersal is 2500BC. I think Ken Nordvedt has arrived at even younger dates.  So, it does make you wonder if the variance methodology needs a bit of recalibration. 


Title: Re: R-P312 update variance numbers by country
Post by: MHammers on January 22, 2011, 11:28:01 PM
The L21 date is about 500 years too young if a beaker model is used.  The latest L21 could have emerged from S116 if is associated with the beaker dispersal is 2500BC. I think Ken Nordvedt has arrived at even younger dates.  So, it does make you wonder if the variance methodology needs a bit of recalibration.  

Most of the dates seem to hover around 2000 BC.  L11+ was probably among the eastern Beaker people, before that either "old Europe" farmers or Yamnaya tribes.  See the Dosideri (sp?) study on Bell Beaker dental traits which explains an earlier western Beaker people (perhaps Hg's E, I, G, J, and T) moving east into south France, Switzerland, and maybe Hungary.  I think if the Yamnaya people were R1b in Hungary at the time they became absorbed by this "Beaker interaction sphere", then spread north and west in the later Beaker movements to Scandinavia and the British Isles via the Rhine.  The intermarriage of these two people may have created the distinct Beaker physical type which was previously unknown.  The origin of Beaker in Iberia in 2900 BC or so may have been the result of the secondary products revolution coming from the east before the main migrations of Yamnaya people around 3100.


Title: Re: R-P312 update variance numbers by country
Post by: Mike Walsh on August 08, 2011, 05:11:23 PM
I decided to look at a by country perspective of R-P312 "All" and get to some of the countries/regions where our sample sizes are not that large. To do that I backed down to accept 37 length haplotypes or greater. Looking at just the 25 non-multi-copy markers of the 1st 37 I came up with the following relative variances.

I was startled a bit to find East/Central Europe (anything east of Germany/Switz/Austria but not the Balkans nor SW Asia) came up higher than France....   so I broke it out as best I could, but it didn't change the ranking....


Hungary_____________:  Var=1.31   (N=15)
Baltic States_______:  Var=1.24   (N=15)
Belarus/Russia/Ukrai:  Var=1.23   (N=26)   
Poland______________:  Var=1.15   (N=26)
France______________:  Var=1.12   (N=188)
Czech Rep.__________:  Var=1.10   (N=12)
England_____________:  Var=1.03   (N=540)
Nordic Area_________:  Var=1.02   (N=71)
Germany_____________:  Var=1.01   (N=181)
Switzerland_________:  Var=0.98   (N=43)
Italy_______________:  Var=0.96   (N=60)
Ireland_____________:  Var=0.94   (N=935)
Wales_______________:  Var=0.93   (N=91)
Iberia______________:  Var=0.92   (N=494)
Benelux Area________:  Var=0.91   (N=43)
Scotland____________:  Var=0.90   (N=463)


Title: Re: R-P312 update variance numbers by country
Post by: MHammers on August 08, 2011, 05:27:16 PM
I decided to look at a by country perspective of R-P312 "All" and get to some of the countries/regions where our sample sizes are not that large. To do that I backed down to accept 37 length haplotypes or greater. Looking at just the 25 non-multi-copy markers of the 1st 37 I came up with the following relative variances.

I was startled a bit to find East/Central Europe (anything east of Germany/Switz/Austria but not the Balkans nor SW Asia) came up higher than France....   so I broke it out as best I could, but it didn't change the ranking....


Hungary_____________:  Var=1.31   (N=15)
Baltic States_______:  Var=1.24   (N=15)
Belarus/Russia/Ukrai:  Var=1.23   (N=26)   
Poland______________:  Var=1.15   (N=26)
France______________:  Var=1.12   (N=188)
Czech Rep.__________:  Var=1.10   (N=12)
England_____________:  Var=1.03   (N=540)
Nordic Area_________:  Var=1.02   (N=71)
Germany_____________:  Var=1.01   (N=181)
Switzerland_________:  Var=0.98   (N=43)
Italy_______________:  Var=0.96   (N=60)
Ireland_____________:  Var=0.94   (N=935)
Wales_______________:  Var=0.93   (N=91)
Iberia______________:  Var=0.92   (N=494)
Benelux Area________:  Var=0.91   (N=43)
Scotland____________:  Var=0.90   (N=463)


Interesting that there is still an east to west cline even at the 37 marker level.  It looks like there was a relatively early jump to England, maybe from the Rhine.


Title: Re: R-P312 update variance numbers by country
Post by: GoldenHind on August 08, 2011, 10:04:41 PM
 Very interesting, but it doesn't surprise me in the least.

As for the Balkans, do we really have a decent sampling there?. If you look at the various P312 subclade maps, there always seems to be a doughnut hole there. Is it because P312 is actually rarish there, or just because we don't have the data?


Title: Re: R-P312 update variance numbers by country
Post by: Mike Walsh on August 09, 2011, 06:40:23 PM
Very interesting, but it doesn't surprise me in the least.

As for the Balkans, do we really have a decent sampling there?. If you look at the various P312 subclade maps, there always seems to be a doughnut hole there. Is it because P312 is actually rarish there, or just because we don't have the data?
We still don't have many R-P312 haplotypes from the Balkans, but I added the south of Italy together with Greece, Croatia, Algeria and Malta to result with this:

Eastern Med Europe__:  Var=0.90  (N=27)   

If I just look at the Balkans, this is all we get

Greece/Croatia______:  Var=0.85  (N=4)

I've never been able to figure out how a Mediterranean route into Europe worked for R-P312, at least when looking at the variance and the phylogenetic trail. It seems like the more samples and resolution, the more indications are that R-P312 moved east to west across the core of the continent.

This not to say that R-L23+ L11- and R-L11* didn't reach Italy and Iberia early, it's just that P312 seems to have come overland and seeped across the Alps into Italy (Cisalpine Gaul) rather than the other way around.


Title: Re: R-P312 update variance numbers by country
Post by: alan trowel hands. on August 14, 2011, 07:39:20 PM
I decided to look at a by country perspective of R-P312 "All" and get to some of the countries/regions where our sample sizes are not that large. To do that I backed down to accept 37 length haplotypes or greater. Looking at just the 25 non-multi-copy markers of the 1st 37 I came up with the following relative variances.

I was startled a bit to find East/Central Europe (anything east of Germany/Switz/Austria but not the Balkans nor SW Asia) came up higher than France....   so I broke it out as best I could, but it didn't change the ranking....


Hungary_____________:  Var=1.31   (N=15)
Baltic States_______:  Var=1.24   (N=15)
Belarus/Russia/Ukrai:  Var=1.23   (N=26)   
Poland______________:  Var=1.15   (N=26)
France______________:  Var=1.12   (N=188)
Czech Rep.__________:  Var=1.10   (N=12)
England_____________:  Var=1.03   (N=540)
Nordic Area_________:  Var=1.02   (N=71)
Germany_____________:  Var=1.01   (N=181)
Switzerland_________:  Var=0.98   (N=43)
Italy_______________:  Var=0.96   (N=60)
Ireland_____________:  Var=0.94   (N=935)
Wales_______________:  Var=0.93   (N=91)
Iberia______________:  Var=0.92   (N=494)
Benelux Area________:  Var=0.91   (N=43)
Scotland____________:  Var=0.90   (N=463)


Very strongly east to west.  The main oddity is France IMO.  If it had been 1.0 or something then the whole pattern would be more logical.  Nevertheless, the general pattern is clear.  It is also true that L21 features that oddity of France having higher variance than Germany.  Certainly for S116 it spoils the otherwise fairly logical variance pattern.