posted
I looked very closely at A3b2, studying all the haplotypes I could find. I was limited to 10 markers, but even then 3 of them showed very clear differences and allowed the samples to be grouped into 4 clusters. These markers are DYS391, DYS385a, DYS385b, and I'm going to refer to their values as 391/385a/385b:
A3b2 Clusters: 1) A European/North African cluster defined by 10/11/12 (A3b2*) 2) An Egyptian/Middle East cluster defined by 10/12/12 (A3b2*) 3) An East African cluster defined by 9/12/12 (A3b2*) 4) An Ethiopia-only cluster defined by 9/12/13 (A3b2b)
A3b2a was found in one person in Sudan, so it's currently considered a private cluster.
The map is very clear. In the Europe/North Africa region most of the samples belong to 10/11/12, which is only found in 1 out of 56 samples in the rest of the world.
The Egyptian/Middle East cluster may be a coincidence, but it would be a big coincidence. It has 3 of 4 samples belonging to 10/12/12, which is found in only 2 of 55 samples from the rest of the world. Plus, there are 4 more samples in this region which haven't been tested for DYS385a/b but do carry the first part of the cluster, DYS391 = 10, which is only found in 2 of 46 samples in East Africa. This makes them likely members of either the Egyptian or European clusters. I found one sample in yhrd.org that came from Turkey and was probably an A3b2 and it had 10/12/12, so to me that indicates the 4 samples between Turkey and north Egypt likely belong to the Egyptian/Middle East cluster, also. But I didn't include the yhrd.org sample because I couldn't be completely sure it wasn't a member of some other haplogroup (namely G2).
[The sample from Cyprus was reported by Capelli. He didn't test for A (M91), but rather found 3 samples that weren't derived for YAP and M89, and assigned them to A after observing their very limited haplotypes. I included 2 of them (Sardinia and Cyprus) because they fit into an A3b2 haplotype, but the 3rd sample, also from Cyprus, was extremely deviant from anything in haplogroup A, and I think it was probably an E1b1b1 that got misreported, which I've seen happen quite a few times in these reports. Anyhow, had it been included as an A3b2, it would have been yet another likely member of the Egyptian/Middle East cluster because it had DYS391 = 10.]
Every sample from north Egypt, North Africa, and Eurasia was SNP-tested as belonging to at least haplogroup A, so it's very unlikely I'm confusing them with some other haplogroup. On the other hand, almost none of the central Egyptian and Ethiopian samples were SNP-tested, but in those places A3b2 is expected to be found and it was easy to detect them.
The Ethiopian A3b2 is split between the 9/12/12 and 9/12/13 samples. Underhill found Ethiopia to be evenly split between A3b2* and A3b2b, and I think they correspond to each haplogroup, respectively. This is backed up by the fact that 9/12/12 is also found in Egypt, Tanzania, and in African Americans, suggesting a more ancient group that diffused farther, while on the other hand A3b2b has so far been only found in Ethiopia (not even in Sudan).
Every one of the 6 samples labeled as 'other' have indications that they are probably members of the local clusters and suffered a mutation in one of their 3 defining STRs. This is especially the case of the Irish sample, which has DYS385a/b = 11/12, which is the typical value of the European cluster, but is seen in only 2 of 51 samples from the rest of the world. The Sardinian 'other' has DYS391 = 11, which almost made me discard it as belonging to A3b2 (it's one of the Capelli 'A' samples), but assuming it's A3b2, that value puts it closer to the European/Egyptian DYS391 = 10 than to the East African/Ethiopian DYS391 = 9, and given that in Sardinia there's another A3b2 which belongs to the European cluster, then it's much more likely this one also belongs to the European cluster instead of the Egyptian. The 4 Ethiopian 'other' samples are more likely to be members of the 2 local clusters than of the Egyptian and European clusters. Every one of the 6 'other' samples can be placed in one of the 4 clusters with just a single step correction in one of its STRs.
In the following map I show the percentages of A3b2 in the world. This map includes all A3b2 samples I found, not just the ones that had an accompanying haplotype. Something interesting to note is that the percentages inside the European and Egyptian regions tend to resemble each other. This is an expected byproduct of the ancient diffusion of these haplogroups. It's kind of like pouring a bag of salt into a pool; it won't remain as a block of salt bouncing around the pool, it will eventually diffuse uniformly throughout the pool. The human species web is the pool. When genetic drift pumps up a haplogroup to a high level, it then starts seeping out to neighboring regions, and becoming uniformly diffused throughout the gene pool. The only problem standing in the way of this explanation is the supposed validity of TMRCA estimates, which say that the great majority of haplogroups coalesced into a single person barely 5,000 to 10,000 years ago.
Judging from the extremely small percentage of A3b2 in African Americans, this haplogroup must be present in West and Central Africa at a rate of less than 0.1%. The western part of West Africa, the Guinean region, has an A3b2 percentage of 0.3%, about 4 times higher then the rest. Together with the fact that 1 of the 2 African American samples belongs to the European cluster, I would say that the Guinean region is the southernmost extension of the European/North African cluster. The other African American sample is a typical member of the East African cluster and it probably comes from the Lake Chad region.
Finally, 2 samples of A3b2 tested for DYS425 and were null for that marker, a very rare value that only occurs in 3% of samples that test for DYS425 (excluding E1b1b samples which are nearly 100% null). It's about the only thing in A3b2's haplotype that stands out as uncommon.
Posts: 39 | Registered: Apr 2007
| IP: Logged |
posted
Not much, but most of the samples available of African origin are from African Americans, so I had to do with what was available. I would have preferred to have thousands of samples from East Africa.
Posts: 39 | Registered: Apr 2007
| IP: Logged |
posted
Obviously, Hg A is relatively less frequent in western Africa, compared to its distribution in the eastern areas. I suspect that the .3% figure was extracted from Rosa et al.'s 2007 study on Guinean populations, but how about citing sources for the figures respectively, for cross-referencing purposes -- because different studies study designated groups that may or may not be the same as that of another study, not to mention sample sizes.
-------------------- The Complete Picture of the Past tells Us what Not to Repeat Posts: 7516 | From: Somewhere on Earth | Registered: Jan 2008
| IP: Logged |
However, the Guinea-Bissau A-M91 lineages do not belong to the widespread A3-M32 but to the A1-M31 subcluster, with reported marginal presence in Mali (2.0% [2,7]), Gambia/Senegal Mandinka (5.1% [7]) and North African Berbers (3.1% [1,33-35]). - Rosa et al.
Posts: 7516 | From: Somewhere on Earth | Registered: Jan 2008
| IP: Logged |
quote:Originally posted by Ausārian: This map ignores this:
However, the Guinea-Bissau A-M91 lineages do not belong to the widespread A3-M32 but to the A1-M31 subcluster, with reported marginal presence in Mali (2.0% [2,7]), Gambia/Senegal Mandinka (5.1% [7]) and North African Berbers (3.1% [1,33-35]). - Rosa et al.
The study also says:
"Any association of Balanta to the Sudanese-speakers is traceable only in the A3b2-M13 and E3* Y chromosomes."
And it has a graph showing what haplogroups they found and it shows they found 1 A3b2 amongst the Balanta.
Posts: 39 | Registered: Apr 2007
| IP: Logged |
posted
I've already pointed out that Hg A is generally less frequent in western African than eastern Africa. The point is however, 1)the thread focuses on only one cluster of Hg A-M91 superclade, and only one cluster of A3b at that,...for what reason(s), I don't know; 2)then the intro map I cited above, purportedly maps out A3b2 frequencies in different regions, but it doesn't produce the source for the said finding in say, for example, shown for the Guinean region -- leaving me to guess and turn to one familiar source that found a low overall frequency of Hg A lineages in a Guinean sample-size of 282 [@ ~ 3.2%], but wherein we are told -- save for the one individual, that the rest of the M91 clusters fell into the Hg A1 clade; Going by that, leaves one with an overall frequency frequency of ~ .4% of A3b2 in the Guinean sample. Cross-referencing from that same source puts A-M91 frequencies in other above-mentioned populations -- i.e. Mali, Gambia/Senegal, Mandinka and North African Berbers -- at 2%, 5.1% and 3.1% respectively, with *no specific reference to subclades*. What's more, is that the frequencies attributed to Gambia/Senegal and North African Berbers are questionable, given that Woods et al. who were referenced for the former, didn't break it down population-wise but language family-wise. In that same study, Hg A3b2 was reported in "Niger-Congo" speaking camp at about .9%, while in Afrasan speaking camp it was about 5.1%, as opposed the highest frequency in the Nilo-Saharan camp of ca. 12.1%. The sister clade of A3b1 was interestingly put at ca. 1.0% in the Niger-Congo camp, while none was reported for either the Afrasan or Nilo-Saharan camps -- the only other group A3b1 was reported for, was the Khoisan, which sported the highest frequency for this clade @ 22%, as opposed to the zip A3b2 reported for them. While as for North African Berbers, Scozzari et al. 1999 didn't exactly use the level of NRY resolution or haplotype conventions as generally used today, which may perhaps be excused for the time it was conducted. The other studies cross-referenced for this group, actually reported no A3b2 [or A-M91 to my knowledge] clades in them. This puts to question, how Rosa et al. generated those figures. Meanwhile, the cited map above, shows A3b2 in the Algerian region at ca. 0.6%, which is obviously inconsistent with Rosa et al.'s cross-references about North African Berbers, while the same map shows A3b2 at 12% in the Ethiopian region for instance, which is seemingly inconsistent with what's been reported in say, for instance, Woods et al., while one might argue that it's closer to the figure reported by Underhill et al. 2000 @ ca. 12.5%...which brings us back to an earlier point:
Obviously, Hg A is relatively less frequent in western Africa, compared to its distribution in the eastern areas. I suspect that the .3% figure was extracted from Rosa et al.'s 2007 study on Guinean populations, but how about citing sources for the figures respectively, for cross-referencing purposes -- because different studies study designated groups [and sub-regions] that may or may not be the same as that of another study, not to mention sample sizes. - by Ausārian
Posts: 7516 | From: Somewhere on Earth | Registered: Jan 2008
| IP: Logged |
All the African countries haven't been tested yet. I'll bet Senegal, Sierra Leone, Liberia and ivory Coast will give similar A frequencies especially within West Atlantic speaking populations.
Posts: 1115 | From: GOD Bless the USA | Registered: May 2006
| IP: Logged |
UBBFriend: Email this page to someone!
Printer-friendly view of this topic