PDA

View Full Version : C-B65 subclade: Closest Relatives of Kostenki-14?



alchemist223
02-07-2021, 05:03 AM
Kostenki-14 (https://en.wikipedia.org/wiki/Kostyonki%E2%80%93Borshchyovo_archaeological_compl ex), a 35,000 year old Paleolithic ancient DNA sample from Voronezh Oblast, Russia, has been added to the YFull tree and has been found to belong to C-Z33130, a new subclade with the sample being positive for the SNPs F9174, Y226190, F8904, Y226191, F8904, F8931, F9348, F9450, and F12012, all of which are shared with the modern-day, rare, subclade of C-B65 exclusively found in China with a TMRCA of 2,800 ybp. In spite of this Y-chromosmal affinity with modern day Chinese ,and per Saag (https://www.ncbi.nlm.nih.gov/pmc/articles/PMC4381518/), men from various Southeast Asian ethnic groups, Kostenki-14 has the mitochondrial DNA haplogroup of U2*, today predominantly found in South Asia as well as in lower levels in Europe and Western Asia. What implications does this have for the peopling of Eurasia? This reminds me of of Oase-1, another Paleolithic-era Eastern European sample (from Romania) with Y-haplogroup K2a*, whose modern-day closest relatives are also found in Eastern Eurasia.

43120 43121

Ebizur
02-07-2021, 06:05 AM
The present-day distribution of C1b1a2-B65 is not really limited to China. That is an illusion created by sampling bias in YFull's data set. It also has been observed in individuals from Singapore (Malay), Borneo (Lebbo from Indonesia and Murut from Brunei), and the Philippines (Aeta).

C1b1a1-SK991/Z16459 subsumes C1b1a1a-M356, which is currently found in South Asia, Southwest Asia, and marginally in Central Asia (e.g. Uyghur), and C1b1a1b-Z16582, which has been observed in a few Arabs in Southwest Asia. Its present-day distribution seems quite congruent with that of Y-DNA haplogroup H(xH2-P96) and K(xK2-M526).

C1b2-B477/Z31885 is frequent in populations of Wallacea and Oceania alongside subclades of K2b-P331(xQR-M45).

It is impossible to make any combination of K+C subclades observe cladality.

alchemist223
02-07-2021, 06:07 AM
I did mention that B65 was present in Southeast Asia, noting the study from Saag, and that Kostenki’s Y-DNA is closer to that of C-B65 than the other subclades of C1b.

Kristiina
02-07-2021, 06:59 AM
One plausible explanation is centrifugal pattern of migrations. Centrifugal pattern is known from linguistics and economics as well. New forms start at the centre and may remain in the peripheral areas. Centrifugal speciation in biology posits that a species' population experiences periods of geographic range expansion followed by shrinking periods, leaving behind small isolated populations on the periphery of the main population.

This is from Dugoujon et al:
GM diversity can be explained by a model of isolation by distance (IBD) at most continental levels, with a particularly significant fit to IBD for the Middle East and Europe. Five peripheral regions of the world (Europe, west and south sub-Saharan Africa, Southeast Asia, and America) exhibit a low level of genetic diversity both within and among populations. By contrast, East and North African, Southwest Asian, and Northeast Asian populations are highly diverse and interconnected genetically by large genetic distances. Therefore, the observed GM variation can be explained by a "centrifugal model" of modern humans peopling history, involving ancient dispersals across a large intercontinental area spanning from East Africa to Northeast Asia, followed by recent migrations in peripheral geographic regions.
https://pubmed.ncbi.nlm.nih.gov/15365983/

davit
02-07-2021, 01:24 PM
Is it possible South Asian C1b and mtdna U2 arrived from a Kosenteki like population?

Pribislav
02-08-2021, 12:29 AM
Kostenki14 was obviously incorrectly assigned to pre-B65 by their software. 3700K BAM of Kostenki14 from Fu et al. 2016 has 145 ancestral and no derived SNPs at B65 level, while shotgun BAM from the original paper (Seguin-Orlando et al. 2014) has only 1 derived SNP at B65 level (Z33130/F12012+ A>T (1T)) and 198 ancestral SNPs. This alone should've caused suspicion if the BAMs were manually checked, but their software just follows the sequence of derived SNPs, while not considering ancestral SNPs at all. The rest of SNPs currently at Z33130 level in YFull's live tree are not covered in Kostenki14, but they were counted as derived, which is another fault in their software that I've pointed out several times in the past. Finally, Kostenki14 has ancestral SNPs at upstream level K281, which means he should in fact split K281 level:


Kostenki14 (3700K BAM; Fu et al. 2016):

K281 level: K380- C>T (11C-1T); B66/Z16458- A>G (5A)


Kostenki14 (Shotgun BAM; Seguin-Orlando et al. 2014):

K281 level: K281+ G>A (1A); K458- T>A (1T); K380- C>T (1C)

alchemist223
02-08-2021, 12:49 AM
Kostenki14 was obviously incorrectly assigned to pre-B65 by their software. 3700K BAM of Kostenki14 from Fu et al. 2016 has 145 ancestral and no derived SNPs at B65 level, while shotgun BAM from the original paper (Seguin-Orlando et al. 2014) has only 1 derived SNP at B65 level (Z33130/F12012+ A>T (1T)) and 198 ancestral SNPs. This alone should've caused suspicion if the BAMs were manually checked, but their software just follows the sequence of derived SNPs, while not considering ancestral SNPs at all. The rest of SNPs currently at Z33130 level in YFull's live tree are not covered in Kostenki14, but they were counted as derived, which is another fault in their software that I've pointed out several times in the past. Finally, Kostenki14 has ancestral SNPs at upstream level K281, which means he should in fact split K281 level:


Kostenki14 (3700K BAM; Fu et al. 2016):

K281 level: K380- C>T (11C-1T); B66/Z16458- A>G (5A)


Kostenki14 (Shotgun BAM; Seguin-Orlando et al. 2014):

K281 level: K281+ G>A (1A); K458- T>A (1T); K380- C>T (1C)

Pribislav, thank you so much for this clarification as well as your SNP analysis. So, once YFull revises their tree, does that mean Kostenki-14 is still a member of a basal C-Z31330 clade, or rather a basal C-K281 clade?

Pribislav
02-08-2021, 01:23 AM
Pribislav, thank you so much for this clarification and your SNP analysis. So, once YFull revises their tree, does that mean Kostenki-14 is still a member of a basal C-Z31330 clade, or rather a basal C-K281 clade?

I don't know where Kostenki14 will be in the next version of YFull tree, but since he has (at least) 3 ancestral SNPs at K281 level it is pretty clear that Z31330+ is a false positive, despite being a transversion. Even K281 assignment wouldn't be particularly strong, since the only derived SNP is a G>A transition covered with only one read. He also has one ancestral SNP at level F1370 (above K281), so I think the safest assignment considering all of the above would be pre-F1370.

F1370 level (merged calls from both BAMs): F1370+ G>C (13C); K384+ G>T (4T); K116+ C>T (4T); Z16480+ T>C (1C); K164+ C>T (1T); Y27076/F19598- C>T (1C)

rock
02-08-2021, 06:17 PM
Z33130 is downstream of B65 as in Karmin 2015 samples Lebb3, Lebb4 from Indonesia are B65+, Z33130-. Dusun8 from Brunei is K281+, B66-. B65+ and M356+ samples are all B66+.

alchemist223
02-13-2021, 10:54 PM
I don't know where Kostenki14 will be in the next version of YFull tree, but since he has (at least) 3 ancestral SNPs at K281 level it is pretty clear that Z31330+ is a false positive, despite being a transversion. Even K281 assignment wouldn't be particularly strong, since the only derived SNP is a G>A transition covered with only one read. He also has one ancestral SNP at level F1370 (above K281), so I think the safest assignment considering all of the above would be pre-F1370.

F1370 level (merged calls from both BAMs): F1370+ G>C (13C); K384+ G>T (4T); K116+ C>T (4T); Z16480+ T>C (1C); K164+ C>T (1T); Y27076/F19598- C>T (1C)

Looks like they've officially incorporated this change into the latest version of their tree. Should we notify them using the above SNP calls?

MNOPSC1b
08-31-2021, 01:40 AM
I suspect the C1b-B65 found among certain Southern Chinese and SE Asian populations are remnants of the Hoabinhian culture. Two other elusive lineages from this region, F2-M427 and D1a1a2-F1070, might also be Hoabinhian-related.

Anyways, I don't think they're particularly close to Kostenki-14, judging by the fact that Kostenki was already West Eurasian, he had already separated from the ancestors of modern East Eurasians more than 40,000 years ago. And there's nothing that indicates he was an upstream of B65.

alchemist223
08-31-2021, 01:22 PM
I suspect the C1b-B65 found among certain Southern Chinese and SE Asian populations are remnants of the Hoabinhian culture. Two other elusive lineages from this region, F2-M427 and D1a1a2-F1070, might also be Hoabinhian-related.

Anyways, I don't think they're particularly close to Kostenki-14, judging by the fact that Kostenki was already West Eurasian, he had already separated from the ancestors of modern East Eurasians more than 40,000 years ago. And there's nothing that indicates he was an upstream of B65.

Yeah, I think YFull flubbed this one, at least according to Pribislav. I remember C-B65 being found in ancient Chinese DNA, but I can't recall the specific study.

MNOPSC1b
08-31-2021, 04:04 PM
Yeah, I think YFull flubbed this one, at least according to Pribislav. I remember C-B65 being found in ancient Chinese DNA, but I can't recall the specific study.

It was found among the native Kradai and Hmong-Mien peoples of ancient Guangxi (Far South of China, at that time hadn't really been sinicized yet) according to Fu Qiaomei et al, with the oldest Guangxi C1b-B65 sample dating back to 1,500 years ago, though the presence of B65 in this region could be much longer, we just haven't found the samples yet. It's very possible that the Laos Hoabinhian sample from 8,000 years ago (currently assigned to C*) might belong to this clade.