Initial Upper Palaeolithic Homo sapiens from Bacho Kiro Cave, Bulgaria[

[Observer]

Pribislav:
32870-32260 BC
"BK-1653 (merged BAMs A11203, A12363, A12562, A12950 and A15722) is a lower coverage sample, but consistent with belonging to haplogroup P-P226, and maybe even pre-R. He is CT+, DE-, C-, F+, F2-, G-, H-, IJ-, L-, T-, K2a-M2335-, K2b1-. This sample is significantly younger than the rest, and only ~2500-3000 years older than Yana samples which belong to clades P-P337 and P-P284. He has a couple of derived SNPs at R and R1 levels, but those are questionable (particularly R1) given his age and ancestral SNPs at P-P226 level. He is also Q- and expectedly R2- and R1a-.

BK-1653; 32870-32260 BC; Bacho Kiro Cave, Dryanovo; Bulgaria; EUP; pre-P-P226 or pre-R

P-PF5867 level: Y507/M1267+ G>A (3A)

P-P337 level: PF5511+ G>A (1A)

P-P284 level: *no calls*

P-P226 level: PF5487/M1201+ G>A (2A); M1215/CTS3813/PF5491- G>C (10G); M1229/F359/PF5920- G>A (2C-1G)

R level: PF6002+ C>T (6T); CTS7876/PF6052+ G>A (1A); PF6034- A>T (1A)

R1 level: PF5474+ G>C (4C)"


YFull:

K2b1Z31091 * P399 * P397+4 SNPsformed 44300 ybp, TMRCA 44300 ybpinfo
K2b1*
M Z30840 * Z30993 * Y27836/Z43403+7 SNPsformed 44300 ybp, TMRCA 10800 ybpinfo
S Z33355 * B254 * Y27620formed 44300 ybp, TMRCA 44300 ybpinfo

P PF5850(H)formed 44300 ybp, TMRCA 41500 ybpinfo
P*
ERS2589709MYS [MY-08] Malaysia
P-M1254Y462/FGC212(H) * PF5853(H) * PF5848(H)+29 SNPs [32SNPs]

P-M1254*
HistAndamanIND [IN-AN]
P-PF5867CTS7481/PF5926/M1234 * PF5855(H) * PF5867+24 SNPs [27 SNPs]

P-PF5867*
P-BY49600BY49600
YF13223PHL Philippines
P-P337PF5901/F1857/PAGES00083/PAGE83/P337 * P243/PF5874 * CTS3697/PF5904/M1212+52 SNPs [55 SNPs]

P-P337*
Yana1RUS [RU-SA]age Yana RHS
P-P284P284 * PF5958/M1160 * PF5482+2 SNPs [4 SNPs]

P-P284*
Yana2RUS [RU-SA]age

Assuming BK-1653 is at this level [though given the age R is possible too], since M1215/CTS3813/PF5491-
P-P226PF5994 * PF5914/F332/M1224 * PF5945/F524/M1248+36 SNPs [39 SNPs]
P-P226*

https://www.yfull.com/chart/tree/P/ [120 SNPs from P* to P226 level]
So about 12000 years min from P* to BK-1653 [depending on how many P226 level SNPs are positive and how many negative]

Bergström et al. 2016 gives slightly earlier divergence time for M & S, both upstream of P

"Among the individuals with indigenous Y chromosomes, 44% belong to haplogroup C, with 42% being C-M347 and 2% the basal C-M130∗. Paragroup K∗ constitutes 56% of indigenous Y chromosomes, with 27% being S-P308, 2% being haplogroup M-M186, and 27% being the basal K-M526∗ "

"Complete sequence data allow direct and accurate inference of the timing of these divergences. Applying a point mutation rate of 0.76 × 10−9 per site per year inferred from the number of missing mutations on the Y chromosome of a ∼45-ky-old radiocarbon-dated Eurasian sample [18], we infer a divergence time of 54.3 ky (95% confidence interval [CI]: 48.0–61.6 ky) between K∗/M chromosomes in Sahul and their closest relatives in the R and Q haplogroups (Figure 1), and a divergence time of 54.1 KY (95% CI: 47.8–61.4 ky) between Sahul C chromosomes and their closest relatives in the C5 haplogroup (Figure 1C), a distinction noted previously on the basis of a single SNP, M347 [9]. These dates are consistent with the archeological record documenting human occupation in Australia by ∼47 kya [2] and with genome-wide analyses that have found an early divergence between the ancestors of Eurasian populations and the ancestors of Aboriginal Australians and Papuans "

[davit]

Bergström et al. 2016 gives slightly earlier divergence time for M & S, both upstream of P

"Among the individuals with indigenous Y chromosomes, 44% belong to haplogroup C, with 42% being C-M347 and 2% the basal C-M130∗. Paragroup K∗ constitutes 56% of indigenous Y chromosomes, with 27% being S-P308, 2% being haplogroup M-M186, and 27% being the basal K-M526∗ "

"Complete sequence data allow direct and accurate inference of the timing of these divergences. Applying a point mutation rate of 0.76 × 10−9 per site per year inferred from the number of missing mutations on the Y chromosome of a ∼45-ky-old radiocarbon-dated Eurasian sample [18], we infer a divergence time of 54.3 ky (95% confidence interval [CI]: 48.0–61.6 ky) between K∗/M chromosomes in Sahul and their closest relatives in the R and Q haplogroups (Figure 1), and a divergence time of 54.1 KY (95% CI: 47.8–61.4 ky) between Sahul C chromosomes and their closest relatives in the C5 haplogroup (Figure 1C), a distinction noted previously on the basis of a single SNP, M347 [9]. These dates are consistent with the archeological record documenting human occupation in Australia by ∼47 kya [2] and with genome-wide analyses that have found an early divergence between the ancestors of Eurasian populations and the ancestors of Aboriginal Australians and Papuans "

Where do you think this divergence took place?

[parasar]

Bergström et al. 2016 gives slightly earlier divergence time for M & S, both upstream of P

"Among the individuals with indigenous Y chromosomes, 44% belong to haplogroup C, with 42% being C-M347 and 2% the basal C-M130∗. Paragroup K∗ constitutes 56% of indigenous Y chromosomes, with 27% being S-P308, 2% being haplogroup M-M186, and 27% being the basal K-M526∗ "

"Complete sequence data allow direct and accurate inference of the timing of these divergences. Applying a point mutation rate of 0.76 × 10−9 per site per year inferred from the number of missing mutations on the Y chromosome of a ∼45-ky-old radiocarbon-dated Eurasian sample [18], we infer a divergence time of 54.3 ky (95% confidence interval [CI]: 48.0–61.6 ky) between K∗/M chromosomes in Sahul and their closest relatives in the R and Q haplogroups (Figure 1), and a divergence time of 54.1 KY (95% CI: 47.8–61.4 ky) between Sahul C chromosomes and their closest relatives in the C5 haplogroup (Figure 1C), a distinction noted previously on the basis of a single SNP, M347 [9]. These dates are consistent with the archeological record documenting human occupation in Australia by ∼47 kya [2] and with genome-wide analyses that have found an early divergence between the ancestors of Eurasian populations and the ancestors of Aboriginal Australians and Papuans "

I would agree that it is older than listed by YFull which IMO underestimates age by ~10-15%.


"Implications for the Peopling of Australia
Y haplogroups from Australia and Papua New Guinea were estimated to diverge from the nearest non-Sahul lineages ∼54 kya, and divergences within Sahul-specific lineages date to ∼48–53 kya. We note that these times post-date the Mount Toba eruption ∼74 kya [22], supporting a model of the initial peopling of this region by modern humans long after this event. The divergence times are close to, but earlier than, the current conservative archaeological date for entry into Sahul, 47 kya [2]. However, the uncertainty in the lineage divergence estimates and the possibility that earlier archaeological sites may be detected make it impossible to determine whether the initial divergence within the Sahul-specific lineages occurred before or after entry into Sahul. The current evidence is consistent with a simple model of a single entry and subsequent rapid lineage divergence."
https://www.ncbi.nlm.nih.gov/pmc/articles/PMC4819516/

[Michalis Moriopoulos]

Well. We have a group that clusters with Ust'Ishim and shows some Asian affinity. Two have mtDNA M, like the Goyet Aurignacians. GoyetQ166-1 has an inexplicable affinity to Han and Tianyuan. Even with the limited use of these samples a piece of the puzzle seems to have fallen into place. Or at least, that is my impression.

It's also useful in putting another constraint on where the Proto-WHG (Tagliente/Grotta Continenza/Villabruna/whatever) clade lived during the days of BK-1653. I know Vestonice and El Miron are modelled as having some Villabruna ancestry so clearly they were in Europe early on; I'm talking about the location of "pure" WHGs before their big expansion. If they were in SE Europe during BK-1653's time (30kya), it must have been south and/or west of the Balkan Mountains, no?

[epoch]

It's also useful in putting another constraint on where the Proto-WHG (Tagliente/Grotta Continenza/Villabruna/whatever) clade lived during the days of BK-1653. I know Vestonice and El Miron are modelled as having some Villabruna ancestry so clearly they were in Europe early on; I'm talking about the location of "pure" WHGs before their big expansion. If they were in SE Europe during BK-1653's time (30kya), it must have been south and/or west of the Balkan Mountains, no?

South of the Balkans is Greece and Anatolia. In Georgia, east of Anatolia there are pre-LGM samples with Basal Eurasian, so East-Anatolia wouldn't likely be the source. Also, from what I understand (West-)Anatolia was uninhabitated during the LGM and the Caucasus as well.
West of the Balkans is an area where most of what is now the Adriatic Sea used to be dry land during LGM.

[K33]

Hey double check the F2 stats you used to run those, do you have maxmiss set to 1?

No, I had the default at maxmiss=0... I thought setting maxmiss to 1 was only ideal when using super-low quality samples?

Since I didn't have the singleton Bacho Kiro Gravettian sample in the f2 run, I thought the 3-sample Bacho_IUP clade was strong enough to run the stats without changing the default maxmiss? I was careful to avoid any lower quality samples like Natufian etc...

Using maxmiss=1 it is indeed harder to generate such tightly-fitting models... but in most of the runs Bacho Kiro_IUP still splits earlier than Tianyuan or Ust Ishim, or at least at the same level as those samples.


EDIT: Below is by far the best model I was able to generate with maxmiss=1, Goyet-Q116 and Taforalt had to be removed from the tree to get the score under 100:




Gravettians modeled as 88% proto-WHG and 12% Bacho Kiro-related, while Yana RHS is 85% of something similar to Ust Ishim (with a little proto-WHG drift) plus 15% Tianyuan-related.

[Kale]

None of those samples are terrible quality, but at maxmiss 0 it's using only the snp's present in 100% of the populations. At maxmiss = 0 even medium quality populations quickly erode the snp count.
The lack of a Neanderthal edge into BachoKiro is unusual, interesting to see the BachoKiro edge into Tianyuan though, that was popping up in models I was running.

[mauors]

out of curiosity I checked F6-620 A11197 for rs1426654, this is one of the two big depigmentation alleles in west Eurasians, basically fixed at A in modern Europeans, he has one call for A
I know from "the genetic history of ice age Europe" that while much of ice age foragers had G as well as most WHGs(Villabruna 5/5 G, Bichon 25/25 G, Loschbour 20/20 G), 1/52 calls for Ust ishim were for A, and 1/6 for Kostenki14 were also A
this is just one call, but I though that was interesting, could be damage too I guess?

BB7-240 A12555 also has one call for G, but that was expected

[aganmu]

I would also add another blow to Basal Eurasian came in the recent Shum Laka DNA paper, which was pubished by the same Harvard geneticists who came up with the Basal Eurasian concept.

They were able to confirm 3 deep-rooted 250-300kya African lineages (including "Ghost Modern", a new discovery) and further clarify the 60-80kya "starburst" event that led to the 3-way split of West/East African and Eurasian lineages-- and yet Basal Eurasian was curiously absent from this whole phylogeny.

Thanks for construe！

[epoch]

Basal Eurasian is an outdated concept Ust_Ishim was instrumental in defining that concept, and then we found Ust_Ishim's brother Oase in Europe, and now I'm sure these BachoKiro folks will join that club.
I think calling it the y-hg IJ branch would be more accurate at this point.
Kostenki14 = 75% Ust_Ishim-like, 25% IJ
Iran_N = 66% IJ, 20% ANE, 10% Excess East-Eurasian, 4% Taforalt-like

What was the extent of area effected by the Campanian Eruption by the way?

Whole of south and middle Italy, Greece. And north of the Black Sea. Kostenki14 was found in a layer right above Campanian tephra deposit. According to the Buran Kaya paper Neanderthals lived in the Krim as Mousterian lithics are found just up to the tephra layers there.

https://www.researchgate.net/figure/...fig7_258612977