World Families Forums - R1b1b2 ht15 cluster in North Africa (385b = 13.2)

Welcome, Guest. Please login or register.
December 25, 2014, 07:38:13 PM
Home Help Search Login Register

+  World Families Forums
|-+  General Forums - Note: You must Be Logged In to post. Anyone can browse.
| |-+  R1b General (Moderator: rms2)
| | |-+  R1b1b2 ht15 cluster in North Africa (385b = 13.2)
« previous next »
Pages: [1] Go Down Print
Author Topic: R1b1b2 ht15 cluster in North Africa (385b = 13.2)  (Read 2093 times)
argiedude
Old Hand
****
Offline Offline

Posts: 146


« on: April 24, 2010, 12:10:21 PM »

About half of North African (Morocco, Algeria, Tunisia) R1b1b2 ht15 samples belong to a cluster that's almost impossible to confuse. It has:

393 = 14
19 = 13
385a/b = 11 / 13.2
389a = 14
389b = 31 (or 389B = 17)

Note that this cluster has nothing to do with another African R1b cluster that also has a .2 mutation in 385a/b. That other cluster is R1b1* and is found almost exclusively in Central Africa.

Amazingly, despite being an R1b1b2 cluster, it's almost impossible to confuse this sample, even if the data doesn't inform of the .2 mutation, such as in ysearch. That's thanks to its 5 different off modal markers on the first 12 STRs.

I've found this sample in almost every study of North African y-dna. It exists in Morocco, Algeria, and Tunisia. I've also found 2 samples of Iberian origin, both of them from Latin Americans. One of the Latin American samples was from smgf, and there weren't any off modal values in the slower STRs. From the 7 or 8 samples I found, it seems this cluster has very low diversity, maybe comparable to Irish M222.

The cluster makes up 0,5% of North Africa's y-dna. About 1% of North Africa's y-dna is R1b1b2 ht15. The 2 samples of Iberian origin constitute 0,01% of Iberia's y-dna (yes, there are more than 10,000 y-dna samples of Iberian origin in ysearch, yhrd, published studies, smgf, etc.).

Another curious thing. In yhrd, I searched for all samples that had 385a/b = 11/13.2 and noticed that, after discarding the obvious members of the North African cluster, there were only 5 samples of R1b1b2 with this mutation, from Iberia, Italy, and North Europe, and all 5 of them had 389a = 14, though they had perfectly typical R1b1b2-ht15 haplotypes, otherwise. That's a hell of a coincidence. In fact, considering that 389a = 14 occurs in only 15% of ht15 samples, that's a 1 in 15,000 coincidence. Weird. Or maybe they are the European relatives from which the North African cluster was born??

I made a map showing the variance of ht15, and there was a very high variance in North Africa, 0,28, higher than anywhere in Europe except Denmark which had 0,29. It was the only result that looked out of place in that map. Now I'm wondering if I included these samples when doing the estimate. If I did, it would easily explain North Africa's high variance, since the samples would be split into 2 equal halves, each with different modals in 5 STRs. But I can't believe I wouldn't have noticed that half of the samples had a .2 mutation in 385a/b. Well, I'll find out later today when I redo that calculation...

BUT HERE'S THE MOST IMPORTANT OBSERVATION. What is this bizarre cluster with an extremely rare .2 mutation in 385a/b? It's haplotype paints it as an indisputable member of R1b1b2-ht15, thanks in large part to the smgf sample from Chile. It's not ht35 (393<>12 and 461=12). But what if it doesn't belong to a typical European haplogroup? I'm thinking specifically of the possibility it's P312- and U106-. And don't tell me that L11+ P312- U106- is typical in Europe. Or perhaps L51+ L11-, but it doesn't look good due to no match whatsoever in its haplotype (L51 has, or tends to have, 426=13, 389B=15, 464d=18, 461=11, and 446=14, and it only matches the North African cluster in 446=14, but this occurs frequently in ht15 despite its modal being 446=13). Anyhow, if it belongs to a P312- U106- haplogroup, this would be... interesting. How would it all fit into understanding how R1b1b2-ht15 diffused through the world? And yes, I think there's a very good chance this cluster could belong to an unexpected haplogroup, because that .2 mutation in 385a/b is not something that occurs every day, it's very, very rare. We have to find one of these guys and have him fully tested. Or maybe write an email to the authors of the 3 or 4 different studies that has detected this sample and explain to them the situation.
« Last Edit: April 24, 2010, 12:12:34 PM by argiedude » Logged

y-dna: R1b L21
mtdna: U5
Maliclavelli
Guru
*****
Offline Offline

Posts: 2158


« Reply #1 on: April 24, 2010, 01:59:08 PM »

I remember having found in the past some individuals with DYS385b=13.2 and different from the African R1b1* above all among British. I don't remember if I put them on Ysearch. Unfortunately my PC has been  put in order again and I lost all my data (I have only my prints, but to find them among thousands and thousands it isn't easy). I confide above all in my memory and if I'll find them I write something.
« Last Edit: April 25, 2010, 09:55:20 AM by Maliclavelli » Logged

Maliclavelli


YDNA: R-S12460


MtDNA: K1a1b1e

Maliclavelli
Guru
*****
Offline Offline

Posts: 2158


« Reply #2 on: April 24, 2010, 02:11:12 PM »

If you find with markers values closer to the modal, you can see that this is an European haplotype from which the North African is derived. Difficult to say from which European country. You know that my heart beats always for Italy.
« Last Edit: April 24, 2010, 02:13:08 PM by Maliclavelli » Logged

Maliclavelli


YDNA: R-S12460


MtDNA: K1a1b1e

Maliclavelli
Guru
*****
Offline Offline

Posts: 2158


« Reply #3 on: April 24, 2010, 03:21:20 PM »

See Ysearch DCQ4V. It is possible  that it is a R1b1b2a1b which had a mutation, a quasi-SNP, on DYS385b from 14 to 13.2.
Logged

Maliclavelli


YDNA: R-S12460


MtDNA: K1a1b1e

Maliclavelli
Guru
*****
Offline Offline

Posts: 2158


« Reply #4 on: April 25, 2010, 03:24:58 AM »

Kitchen (UX5UR) and Jones (7NY99) have also DYS385b=14.2. It is then clear that the European variance is higher than the North African one.
Logged

Maliclavelli


YDNA: R-S12460


MtDNA: K1a1b1e

Maliclavelli
Guru
*****
Offline Offline

Posts: 2158


« Reply #5 on: April 25, 2010, 04:15:51 AM »

Anyway the closest to the North African haplotype is an Italian from Brescia (from Celt Brixia, but in the North of the Rhaetian Etruskan Fatherland):

DYS19= 14 (13)
DYS389I-II=14-31
DYS390=24
DYS391=10 (11)
DYS392=13
DYS393=13 (14)
DYS385= 11-13.2
(from YHRD)
« Last Edit: April 25, 2010, 04:16:20 AM by Maliclavelli » Logged

Maliclavelli


YDNA: R-S12460


MtDNA: K1a1b1e

argiedude
Old Hand
****
Offline Offline

Posts: 146


« Reply #6 on: April 25, 2010, 03:31:28 PM »

All percentages are over the total y-dna:

__________ht15 (13.2)___ht15 (all)
Morocco_____0,5%_______2,5%
Algeria______0,8%_______7,0%
Tunisia______1,2%_______2,5%

Algeria has only 120 samples, so the big percentage of 'ht15 (all)' could be just a statistical anomaly.

I said that the 13.2 cluster made up half of North African ht15 samples, but I was wrong about the frequency of ht15 in North Africa, which is more like 2%, not 1%, so the cluster is about one quarter of all North African ht15 samples. And notably homogenously spread out from Morocco to Tunisia.

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

The variance I calculated for North African ht15 was indeed flawed. I didn't notice there were so many samples with a .2 mutation because the columns in Excel cut them off. Now that I redid the calculations, the 13.2 cluster has a variance of 0,07, extremely low, probably less than the Irish M222+ clade. And the rest of ht15 have 0,24, dropping from the previous estimate of 0,28 which included the 13.2 samples. 0,24 is still not a low figure, it's the same as Iberia and slightly higher than Italy. And slightly lower than North Europe.
Logged

y-dna: R1b L21
mtdna: U5
Maliclavelli
Guru
*****
Offline Offline

Posts: 2158


« Reply #7 on: April 25, 2010, 04:07:13 PM »

If you have wide marker values, you can calculate more deeply the origin of these haplotypes. I put to your disposal some individuals I found on SMGF. Unfortunately for to-day I exhausted my time to research, because there are other individuals on SMGF from some European countries.

Remember that North Africa was many centuries under the Roman Empire, with colonies, some certainly Etruskan, and there is always the Vandals among those peoples.

A variance of 0,07, less than R-M222, probably presupposes a recent introgression. I wouldn't exclude a Jewish origin, whichever is then the real origin of those Jews. Some Americans strangely have Jewish names and come from  Jewish named towns, even though they could come from some Christian church.
Logged

Maliclavelli


YDNA: R-S12460


MtDNA: K1a1b1e

argiedude
Old Hand
****
Offline Offline

Posts: 146


« Reply #8 on: May 02, 2010, 01:04:58 PM »

Well! I got some exciting news.



If the above image is very small, then click on this link to view it in full size:
http://i88.photobucket.com/albums/k178/argiedude/R1b1b2-cl385b13.gif

Here's a link to the excel file with the all data about this cluster:
http://www.sendspace.com/file/524n55

Ok. I found several samples of the European cluster of R1b1b2-385b=13.2, and a few of them were uploaded to ysearch, which lists their haplogroup. 3 of them had unknown or  R1b listed as their haplogroup, but one listed itself as... R1b1b2h, aka U152. This sample has all the hallmarks of this cluster. It has 385b=13.2 (not 14 or 14.2). It has 389a=14 (not 13). It has 389B=17 (not 16). The vast majority of R1b1b2 samples with 385b=13.2 have 389a=14 and 389B=17, it's an obvious cluster. This sample that listed itself as U152 has all these characteristics. There's no doubt it belongs to this cluster of 385b=13.2. Ergo, barring the possibility he mislabeled himself (which is indeed a real though slight possibility in ysearch), we can be pretty certain that this cluster of R1b1b2 belongs to U152!

Remember there's actually 2 clusters here, a European cluster (385b=13.2, 389a=14, 389B=17 [aka 389b=31]), and a North African cluster, which has everything the European cluster has, plus 393=14 and DYS19=13. Both have very low variance, with the European cluster having 0,13 and the North African 0,08. Under standard operating procedure of TMRCA theory, this would indicate an age around 2,000 years ago.

Some curious "coincidences". The North African cluster is generally homogeneously diffused across North Africa, but it has a slight east-west gradient, with Tunisia being the place with the highest percentage (1%), and Morocco the lowest (0,5%). Of course, U152 is much heavier in Italy than it is in Iberia, so it all seems to kind of fit. But the European cluster isn't centered in Italy. Instead, it seems to be yet another case of a cluster that is exceptionally rare (in example, less than 1 per 1,000 samples) and rather homogenously distributed across all of West Europe.

In the file there are 4 ysearch samples that are definitely members of the European cluster, because their sample was listed with a .2 mutation in smgf, and 2 ysearch samples that are U152 and are very close to the European modal but we don't know if they do or don't have a .2 mutation in 385b. We should contact these known and potential members of this cluster to tell them about this, and of course, try to get them to fully test themselves for all downstream U152 SNPs.

So... there's a North African cluster of R1b1b2, which makes up 1/4 of their R1b1b2 samples... and it's U152. :)

EDIT: In the variance map of ht15 that I made in another thread, about the only region that had a value that didn't quite "fit" was North Africa, which had one of the highest variances of ht15 anywhere in West Asia: 0,28. Now we can see that this was an artifact of the very unusual modal of this cluster, which makes up 1/4 of their R1b1b2 samples. Excluding this cluster, the remaining North African R1b1b2 have a variance of 0,24, identical to Iberia, and slightly higher than Italy, which has 0,22. North Africa's R1b1b2 has an east-west gradient, with higher percentages in the west, even more so if we exclude this North African cluster, which has higher percentages in the east. So it seems pretty clear to me that the picture consists of a 2-way diffusion, mostly from Iberia, but also significantly from Italy (the Italian diffusion consisting basically of just the .2 cluster).

EDIT II: Another likely modal value is 456=15. It's usually 16 in R1b1b2. It's a fast marker, though. The ysearch sample that listed itself as U152 has 456=14, reenforcing the fact that he is a member of this .2 cluster.
« Last Edit: May 02, 2010, 01:23:51 PM by argiedude » Logged

y-dna: R1b L21
mtdna: U5
Maliclavelli
Guru
*****
Offline Offline

Posts: 2158


« Reply #9 on: May 02, 2010, 01:14:58 PM »

Great work, Argiedude!
Logged

Maliclavelli


YDNA: R-S12460


MtDNA: K1a1b1e

Nolan Admin - Glenn Allen Nolen
Project Coordinator
Old Hand
*****
Offline Offline

Posts: 292


WWW
« Reply #10 on: May 02, 2010, 04:04:52 PM »

“Great work, Argiedude!”

I second that! Kudos, you appear to be doing great work in this field, I for one thank-you.
Logged
argiedude
Old Hand
****
Offline Offline

Posts: 146


« Reply #11 on: May 02, 2010, 04:15:30 PM »

Here's an updated version of my map of ht15 variance.



If the above image is very small, then click this link to see it in full size.
http://i88.photobucket.com/albums/k178/argiedude/R1b1b2variance-ht15updatedaftery-1.gif

North Africa dropped from 0,28 to 0,23. Its former high value was because of the 385b=13.2 cluster. By the way, the exclusion of clusters that have a different modal haplotype is completely warranted. They always increase the variance, even if their own internal variance is very small. It's an artifact of their different modal haplotype. On the other hand, including different haplogroups such as U152, SRY2627, or L21 isn't a problem, because their modal haplotypes are the same. The only thing that's left me wondering is wether the apparent east-west cline, higher in the east, might be due to a higher proportion of eastern R1b1b2 belonging to U106, because U106 has one STR off modal value, but it's one of the STRs I'm using to calculate variance (it has 393=23 instead of 393=24). But I think Tim Janzen at rootsweb, whose been making variance estimates of individual ht15 haplogroups, is also getting a slightly higher variance in the east (northeast, really), not just in U106.

EDIT:

Great work, Argiedude!

“Great work, Argiedude!”

I second that! Kudos, you appear to be doing great work in this field, I for one thank-you.


:)
« Last Edit: May 02, 2010, 04:16:55 PM by argiedude » Logged

y-dna: R1b L21
mtdna: U5
Maliclavelli
Guru
*****
Offline Offline

Posts: 2158


« Reply #12 on: May 03, 2010, 09:09:16 AM »

As also you noted, about the sample from Reunion, probably we must reflect a little bit about his belonging.
We can suppose he is of French, then European, origin: but he has not only DYS393=12 instead of 13, but also DYS385a=11.2: the unique thus far. We should presuppose that a kind of RECHLOH was happen in a way we don't know.
Perhaps it is more interesting to think he belongs to a different cluster, an African one, we don't know which haplogroup or clade it belongs to.

Re. your hypothesis that this clade is R-U152, it is based uniquely on the belonging of Bricker to this haplogroup.
But Bricker is a friend and you can write to him asking if he is really DYS385b=14.2. But the hypothesis is that DYS385=13.2 has been a mutation from 14, then Bricker could belong to the cluster from which this one derived.

I hypothesized that this haplogroup was an ancient R1b1b2a1b gone out of the Italian Refugium many thousands of years ago and not two thousands.
If it was as you are hypothesizing, we should think to a Roman, who, like Genghis Khan, has spread his Y in so vast space.
Should we think to Julius Caesar, who, as Romans said, had gone bed not only with all the men of Rome, but with all the women?
Logged

Maliclavelli


YDNA: R-S12460


MtDNA: K1a1b1e

Maliclavelli
Guru
*****
Offline Offline

Posts: 2158


« Reply #13 on: May 03, 2010, 09:43:06 AM »

And why to think to the North African haplotype present in Spain or the European one present in North Africa as to a recent introgression and not to an ancient presence? Then to a more ancientness of both?
Logged

Maliclavelli


YDNA: R-S12460


MtDNA: K1a1b1e

MHammers
Old Hand
****
Offline Offline

Posts: 347


« Reply #14 on: May 03, 2010, 12:21:24 PM »

Nice map Argiedude,

What markers are you using to get the variance?  I think the east is generally higher as well.  I have .38 for Ukraine which is probably too high due to a small sample, but my Russian sample of .28 for L51+ types is comparable with your map.

I used DYS19, 390,391,392,393, 389i-ii, 385 a/b from YHRD. 

Logged

Ydna: R1b-Z253**


Mtdna: T

argiedude
Old Hand
****
Offline Offline

Posts: 146


« Reply #15 on: May 03, 2010, 06:21:25 PM »

Nice map Argiedude,

What markers are you using to get the variance?  I think the east is generally higher as well.  I have .38 for Ukraine which is probably too high due to a small sample, but my Russian sample of .28 for L51+ types is comparable with your map.

I used DYS19, 390,391,392,393, 389i-ii, 385 a/b from YHRD.  

Mike, to make the maps of ht15 and ht35 variance (I'll try to refer to them with the current nomenclature in the future), I used 6 STRs: 390, 19, 391, 389a, 392, 389B (= 389b - 389a).

I didn't use 393 because I was predicting the samples from the yhrd darabase, so as to get a good coverage of southeast Europe, because of the hugely lopsided northwest European concentration of samples in commercial databases, so 393=12 was ht35 and 393=13 was ht15 (I'll try to use current nomenclature in the future). And thus, it wouldn't make sense to use 393 as part of the variance estimate.

I also didn't use 385a/b because of the problems presented by these types of STRs, what's their name again, multi-copy STRs, I think. They tend to produce higher variances than would be expected from their known mutation rate, because of the additional factor that they occasionally suffer RecLoh copy-overs.

Did you use ysearch samples for estimating Russian and Ukrainian variance? Most ysearch samples from Russia, Ukraine, Lithuania, Belarus, etc. are Jewish, and they have a notable tendency to form their own clusters, even in L21, with their own particular haplotypes and extremely reduced variance, that are not representative of the country.
« Last Edit: May 03, 2010, 06:24:57 PM by argiedude » Logged

y-dna: R1b L21
mtdna: U5
MHammers
Old Hand
****
Offline Offline

Posts: 347


« Reply #16 on: May 03, 2010, 08:00:26 PM »

No, I didn't use ysearch which is something I will probably look at in the future.  

I agree about the Jewish populations in the east, which are problematic for interpretation of origin and diffusion.  We still don't know what percentage are descendants of indigenous converts vs. 1st millenium diaspora people from SW Asia/Levant.  Both sources had R1b1b2 long before Judaism.  

I included 385 in my final calculations and am aware about RecLoh events skewing the variance, but I also removed it just to see if I still had higher variances from those locations.  Since I wasn't calculating age estimates, I didn't think it was too important.  They were still higher, at least, in my L51- variances.  L51+ may be a different story.  

I used 393 in YHRD to define L51-/L51+.  11 and 12 for L51- and 13 and 14 for L51+.  This is not entirely reliable, but I think it supports the general picture of things SW Asia=older, Central/West Europe=younger.  It's these intermediate areas of eastern Europe, the Caucasus, and SE Europe including Italy, that unfortunately, are not as clear for L51-.  L51+ does seem to go east to west more clearly.
« Last Edit: May 03, 2010, 08:07:43 PM by MHammers » Logged

Ydna: R1b-Z253**


Mtdna: T

Pages: [1] Go Up Print 
« previous next »
Jump to:  


SEO light theme by © Mustang forums. Powered by SMF 1.1.13 | SMF © 2006-2011, Simple Machines LLC

Page created in 0.108 seconds with 18 queries.