Relative Proportions of R1b1b2 Groups in Ysearch

[argiedude]

I did this map to approximate in some way the frequency of L21 in different regions. We all know how difficult it is to gauge L21's presence because of the distortions in how many people get tested per region, the different dates at which each SNP became available, and the non-reporting of negative results in the ysearch database. What I did was to take the number of L21 samples per region found in ysearch, and then divide it by the number of total samples from that region in ysearch. This won't give us an estimate of the absolute frequency of L21 in each region, but it will give us relative proportions of the frequency for each region, which realistically is more valuable than the absolute frequency itself.

For example. There are 6400 samples from Ireland (in ysearch). And there are 102 L21+ samples from Ireland (in ysearch). This is a frequency of 0,0159, and that's the number I put for Ireland in the map: 159. This doesn't mean that L21+ makes up 0,0159 (1,5%) of Irish y-dna; the actual frequency of L21+ in Ireland, excluding M222+, is probably 50% or more. But it allows us to compare regions with each other, and start to paint a picture of where it's relatively more common.



Here are 5 tentative observations of my own:

1) There are 2 exceptional hotspots: Scotland/Ireland/Wales and Norway. Norway might be a statistical fluke, since it's composed of just 5 samples, but Ireland/Scotland are not an anomaly. The situation for Wales is a fence-sitter.

2) The rest of West Europe has a homogenous distribution. England's low result (62) is confirmed by the USA and Unknown results; I have confirmed repeatedly over the last year that the results labeled USA and Unknown are overwhelmingly of English/Wales origin, because of their frequencies of R1b, R1a, M222, and other haplogroups. I estimate the origin of these samples as 85% English, 10% Scottish/Irish, and maybe 5% other (German, African American, etc.). The USA and Unknown samples number in the tens of thousands. Their validity is the best. And they are saying that England's rate of L21 (49 to 60) is in the same range as West Europe (30 to 70).

3) We can't make precise country to country comparisons. One thing is Scotland and Ireland who are way off the charts, but we can't conclude that France's rate of L21 (65) is higher than Germany's (36). The situation could be the exact opposite. We can only make broad sweeps of judgement, at this stage. As an example, my own sample from north Italy has pumped up the rate of this region to 69, which compared with Ireland, where L21 is probably half of the y-dna, would suggest north Italy's frequency of L21 is 20%. But thanks to the results of the North Italy Project, we know that the rate of L21 in north Italy is very likely going to range between 1% and 5%. It's just a statistical anomaly, we have too few results. But they are very similar across West Europe, including England, so on that broad count alone I think we can reach a conclusion. But again, we can't (yet) make specific country to country comparisons.

4) East Europe seems to be seriously lacking in L21+. I've included 2 samples, in blue, from Denmark and Croatia, reported by Steve yesterday. They're not in ysearch, so I put them in blue. To arrive at the Croatian result I combined all neighboring southeast European regions. So the very low figure shown is the rate of L21+ in southeast Europe, not in Croatia. On the other hand, for Denmark it's the rate for that country. I know there are 2 samples from Lithuania and Romania, but given that they are from a German Romanian and an Ashkenazi, their origin is obviously more likely to be West Europe than Lithuania and Romania. Now that I think about it, I have a feeling I just presumed the Lithuanian L21 is from an Ashkenazi. I usually assume that ysearch samples from the Baltic countries and Belarus are almost exclusively from Ashkenazi Americans.

5) I'm going to guess the absolute frequency of L21* in continental West Europe, including England. Given that the figures uniformly hover around 50 +/- 20, and that Ireland has a rate of 159, and that Ireland probably has an absolute frequency of L21* (no M222+) of 50% (over their total y-dna), I'd say that the frequency of L21* in continental West Europe + England is 30% (over their total y-dna). [edit next day: I miscalculated. The percentage of L21 in continental Europe plus England would be 15%, not 30%]

PS: You might think it's crazy that Finland or North Italy could have a rate of L21 comparable to England. I estimate there are just 150 to 200 north Italian samples in ysearch. There are 50 times as many from England (about 8,500 samples). There's 1 L21+ in ysearch, and there are 56 L21+ from England. There you go.

PSII: Is it normal for the image too appear that small? [in the meantime, right click and select view image]

[rms2]

Nice work, argiedude, but I don't get "View Image" when I right click on it.

Advice?

[argiedude]

Hmmm, you must be using a different browser. Has this always been an issue? Seems like something that has to be adressed.

[GoldenHind]

Hmmm, you must be using a different browser. Has this always been an issue? Seems like something that has to be adressed.

I don't get the view image option either.
I suspect that the precentage of L21 in Norway will turn out to be greater than that of Finland.

[argiedude]

This is insane. The image is small because the image tag has this:

[img width=85 height=78]

instead of this:

[img]

So, I edit and change the tag back to normal:

[img]

and when I finish the edit and save it, it automatically reverts to:

[img width=85 height=78]

That's what's causing the problem. I edited it 5 times, and it automatically reverts back.

[Mikewww]

I did this map to approximate in some way the frequency of L21 in different regions. ...  3) We can't make precise country to country comparisons.....

Great work, argiedue!

P.S. I just view or download and then copy it into Powerpoint and do whatever I want to with it from there.  Firefox makes this easy.

[cmblandford]

Or right click and convert to an adobe pdf; adjustment can be made for size and the image is very clear.

[Nolan Admin - Glenn Allen Nolen]

Right Click: Save Picture to your computer

[vtilroe]

Nice work.  I suspect as deep-clade upgrades continue over the next few years, the continental aspect of L21 will begin to resolve a little better.  I do however, find the present poor showing in the Low Countries rather odd in context with Germany, Scandinavia, etc.

[Jafety R1b-U152]

I also find the Low countries low rate interesting. However, there is a "hotspot" of R-U106. I can thus imagine two different versions:

1. U106 was there on the lower Rhine when L21 migrated out of the Franco-German broder area. They passed U106 by and so reached the British Isles and Scandinavia.

2. Most of Holland could have been under sea thus uninhabitated during the L21 migration, and was only populated when non-L21 population became dominant in the region.

For Eastern Europe, I would be surprised if someone would find L21 (or U-152) in West or East Slavic speaking areas (who are not of Jewish or German ancestry), where R1b1b2 is relatively low and most of them U-106 from East Germanic peoples. In Hungary and the Balkans however, there could be some as a sign of Celtic migrations.

[argiedude]

Another one! :)

This one is U152+ (including all subclades)



Steve, could you change the title of this thread to Relative proportions of R1b1b2 groups in ysearch ? There are more coming! :))

[Nolan Admin - Glenn Allen Nolen]

Very nice work! Thanks, I appreciate the inclusion of Ireland in your R-U152 Y-Search frequency map. I'm actually proud to be part of this industry, again.

Thanks for your hard work!

[argiedude]

And there's even more. :)



I have to go back and make a correction to the U152 map. North and south Italy will appear with more U152.

edit: done. North Italy seems to have everything in R1b1b2.

edit 2:
Yesterday I made a rough estimate that L21 would be 30% of continental Europe's y-dna. I made a miscalculation, so the actual estimate would be 15%, not 30%. That is, 15% of the y-dna of continental Europe and England should be L21+. We'll see...

[argiedude]

And finally, SRY2627

[argiedude]

P312*



, the numbers don't seem to square for England.

L21.....20%
U109..13%
U152..  2%
SRY2627...1%
P312*...don't know (5%?)

Total....41%

England's R1b1b2 is 60% to 65%.

[Jafety R1b-U152]

Very nice maps, thanks.

If we look at these, Northern Italy or France seem a more likely origin of P-312 than the Rhine valley. Interesting. SRY2627 is basically M167, am I right?

[rms2]

Nice work.  I suspect as deep-clade upgrades continue over the next few years, the continental aspect of L21 will begin to resolve a little better.  I do however, find the present poor showing in the Low Countries rather odd in context with Germany, Scandinavia, etc.

I was looking at the Benelux Project this morning. My first impression is that a very very low proportion of the R1b1b2s in that project have had any kind of SNP testing. Of course, that doesn't change the fact that of those who have been SNP tested recently, by far most of them have been L21-. But, still, the sample size is extremely small, especially when compared with the number of un-SNP-tested R1b1b2s in that project.

I found a guy a few days ago who it seemed to me is likely to be L21+. The surname was ten Eyck. I emailed him and asked him about testing for L21, but he never answered. Of course, that is not at all unusual. That happens with most of those I try to recruit from YSearch, unfortunately.

[Maliclavelli]

Many thanks, Argiedude. I think that your work is the best gift (and demonstration) of the theories that my friend Gioiello Tognoni has supported in these last years. The different percentages of South Italy are due, as Gioiello noted many times, by the neolithic introgression from South Balkans, which diluted, but didn't change, this reading of the data, which is also the yours.
Also the theory of an ancient presence in Italy (and in Europe) of J1 is now demonstrated by a paper of Tofanelli (someone  would argue that also Tofanelli as Gioiello is a Pisan: but we have now the return of the prodigal son Cristian Capelli, and he isn't from Pisa and is teaching in one of the most prestigious universities of the United Kingdom).

[Jafety R1b-U152]

Could you send a link to that paper, please? I thought J1 came during the Arabic times to Italy, the new idea is quite interesting.

[Maliclavelli]

Dear Jafety, you can find the paper of Tofanelli et alii on "Dienekes' blog". Unfortunately it isn't for free. This is the abstract:
"The present day distribution of Y chromosomes bearing the haplogroup J1 M267*G variant has been associated with different episodes of human demographic history, the main one being the diffusion of Islam since the Early Middle Ages. To better understand the modes and timing of J1 dispersals, we reconstructed the genealogical relationships among 282 M267*G chromosomes from 29 populations typed at 20 YSTRs and 6 SNPs. Phylogenetic analyses depicted a new genetic background consistent with climate-driven demographic dynamics occurring during two key phases of human pre-history: (1) the spatial expansion of hunter gatherers in response to the end of the late Pleistocene cooling phases and (2) the displacement of groups of foragers/herders following the mid-Holocene rainfall retreats across the Sahara and Arabia. Furthermore, J1 STR motifs previously used to trace Arab or Jewish ancestries were shown unsuitable as diagnostic markers for ethnicity".

[Jafety R1b-U152]

Argiedude, could you make a similar "relative Y" map for M269+, U106-, P312- people? That would be also interesting, especially if Northern Italy would be showing strong numbers...

[rms2]

The trouble with YSearch is that the subclade reporting is not accurate. There are some people who are all tested up who have never updated their haplogroup entries.

I guess if you do anything with YSearch you just have to keep that in mind and accept that there will be a fairly large margin of error.

[Jdean]

The trouble with YSearch is that the subclade reporting is not accurate. There are some people who are all tested up who have never updated their haplogroup entries.

I guess if you do anything with YSearch you just have to keep that in mind and accept that there will be a fairly large margin of error.

I did a similar thing to this a little while back, admittedly without the nice graphics, using a WMHA member of the Stedman project as the bases, and ended up with very similar results, I can't quite remember how I did at now though.

I agree with Y-Search haplogroups being iffy though, some are reported when they haven't been tested for, and there is also a certain amount of human error in entering them, presumably not helped by the names changing all the time.

[rms2]

The trouble with YSearch is that the subclade reporting is not accurate. There are some people who are all tested up who have never updated their haplogroup entries.

I guess if you do anything with YSearch you just have to keep that in mind and accept that there will be a fairly large margin of error.

I did a similar thing to this a little while back, admittedly without the nice graphics, using a WMHA member of the Stedman project as the bases, and ended up with very similar results, I can't quite remember how I did at now though.

I agree with Y-Search haplogroups being iffy though, some are reported when they haven't been tested for, and there is also a certain amount of human error in entering them, presumably not helped by the names changing all the time.

What I am finding, too, is that R1b1b2 haplotypes of different subclades are remarkably, frustratingly close. I go through YSearch all the time running the haplotypes of various members of the R-L21 Plus Project and then trying to recruit their haplotype neighbors for Deep Clade-R testing and, if they turn out to be L21+, for project membership.

I used to think a 33/37 "match" was a fairly good one. Not anymore. There are men of different subclades that are that close at 37 markers! Just today I sent a recruitment email to one in particular, a 33/37 "match" to one of our members. That person answered me saying the test had already been done and came up U106+ L48+! That's not even P312+, let alone L21+! (Now you see why I am putting quotation marks around the word match.)

I can only conclude that R1b1b2 (or maybe R1b1b2a1) must have expanded very rapidly, fairly recently. What else can account for all the similarity in haplotypes between different subclades of R1b1b2?

It makes it tough to spot an L21+ haplotype, that's for sure.

[Jdean]

What I am finding, too, is that R1b1b2 haplotypes of different subclades are remarkably, frustratingly close. I go through YSearch all the time running the haplotypes of various members of the R-L21 Plus Project and then trying to recruit their haplotype neighbors for Deep Clade-R testing and, if they turn out to be L21+, for project membership.

I used to think a 33/37 "match" was a fairly good one. Not anymore. There are men of different subclades that are that close at 37 markers! Just today I sent a recruitment email to one in particular, a 33/37 "match" to one of our members. That person answered me saying the test had already been done and came up U106+ L48+! That's not even P312+, let alone L21+! (Now you see why I am putting quotation marks around the word match.)

I can only conclude that R1b1b2 (or maybe R1b1b2a1) must have expanded very rapidly, fairly recently. What else can account for all the similarity in haplotypes between different subclades of R1b1b2?

It makes it tough to spot an L21+ haplotype, that's for sure.

A very resent search of Y-Search based on my own slow markers produced 21 results one of which was R1b1b2a1, a Stevens as it happens, and no others tested past R1b1. I checked the Stevens one just in case and it was right.

A similar search on another member of the Stedman project came up with 17 results 7 were tested  R1b1b2a1 with no others past R1b1, it's the luck of the draw I suppose.