(12-13-2024, 06:39 PM)crashdoc Wrote: C1b and C1a could already have been present in the hub before the Eurasian dispersion event. TMRCAs on yfull and ftdna are too young. For instance now that we have 2 high coverage males dating from around 45kya (Ranis13 & Ust-Ishim), using their private mutations combined with their radiocarbon dating we can calculate with quite a high reliability that the real TMRCA of K-M2308 is between 48-49kya, while y-full puts it's parent (K2) at 45.5kya and ftdna at 45kya. The dispersion event cannot have happened before then as we have Ranis, Ust-Ishim, NO and South-East-Asian clads under that.
Hallast et al. (2023) have presented split times as estimated according to BEAST analysis of fully sequence-resolved Y-chromosomes in their Figure S1:
K2 (MRCA of NO and QR Y-chromosomes): TMRCA 52,000 (95% HPD interval 45,500 <-> 59,600) ybp
NO (MRCA of N and O Y-chromosomes): TMRCA 45,300 (95% HPD interval 39,400 <-> 51,900) ybp
For whatever reason, many people seem to ignore the depth of the split between Y-DNA haplogroup N-M231 and Y-DNA haplogroup O-M175. The K2a lineages found in Paleolithic specimens from Europe are basal to the MRCA of N-M231 and O-M175, so they are indeed highly divergent.
The data of the IVPP (the Institute of Vertebrate Paleontology and Paleoanthropology) imply that the origin of the differing MRCAs, obtained by Hallast already after the actual splits between lineages had already been inferred in other earlier works, should not be related to the origin and distribution of the bearers of yDNA K population, including bearers of yDNA K2, yDNA O-M175 and yDNA N-M231.
Regarding the above data, which were posted by Ebizur previously, it should not be thought that the data of the IVPP did not address such ages, therefore, it should not be thought that these ages were necessarily related to actual accurate consequent proto-historical splits between y chromosome lineages, cited above.
For example, the cited age of yDNA CT, such as 77800 years ago (the “split” between African and Eurasian lineages, listed above), is older than the age of 76500 years ago, which was a split between yDNA CF and yDNA DE in “A Rare Deep-Rooting D0 African Y-Chromosomal Haplogroup and Its Implications for the Expansion of Modern Humans Out of Africa”, but this age of 77800 years ago can still be an age of one of yDNA CT* populations, some Homo Sapiens remains of which had previously coexisted to an insignificant degree in the neighbourhood of other groups of ancient Homo Sapiens individuals in accordance with the IVPP data. The age of 77800 years ago “resurfaced” for the DNA from Australia. Which language could this type of yDNA CT* people speak? For example, Gerhard Jager cited an example of the convergent evolution in the English language and one of languages of the Australian aboriginals in the article “From words to features to trees: Computing a world tree of languages from word lists”:
Old English docga >English dog Proto-Paman *gudaga >Mbabaram dog ('dog')
The IVPP data to a certain degree defended such a 77800-year-old yDNA CT* population in Eurasia and did not suggest the association of this population with the population, characterized by mutation C16355T, observed in Homo Heidelbergensis.
On the contrary, the IVPP data implied that some remains of this 77800-year-old yDNA CT* population in Eurasia became incorporated in the Homo Sapiens component, which is not usually associated with a source of archaic admixture.
eurasia Wrote:We have at least three Denisovan introgression events in East Asia.
1) Tibetan introgression, with EPAS1, closer to Altai Denisavan (Tibetans have >=2 introgressions, I'm referring to EPAS1-related only)
2) Tianyuan-Mainland East Asia (Denisovan ancestry and population history of early East Asians), without EPAS1, distant to Altai Denisovan (The history and evolution of the Denisovan-EPAS1 haplotype in Tibetans)
3) Papuan-Australian, also distant to Altai Denisovan
and others need further investigation to separate from the ones above
4) a recent Siberian one which lacks EPAS1 but close to AD.
5) Aeta-Agta, higher than Papuan
6) India
Even if Tianyuan has some Altai Denisovan-related introgression, the amount would be small.
Makes sense that there would be separate Denisovan introgressions among Papuans-Australian aborigines, Tibetans, Philippine Negritos, General East Asians (that should also be found among Tibetans and Philippine Negritos at least), and AASI.
Here's a graph attempt just to get someone to post a better one. I think I extracted f2 with maxmiss=0.1 and there were around 600k SNPs remaining. Worst residuals are bad but I can't see a straightforward modification from them. Added Denisova and Neanderthal to see if they help, graph works the same way without them.
from to type weight low high
Root_Rootf Rootam edge 0.04187979161912129 NA NA
Root_Rootf Root_Rootf_Denisova.DG edge 0.04187979161919381 NA NA
Rootam Roota edge 0.04545448296559132 NA NA
Rootam Mbuti.DG edge 0.010979429137138976 NA NA
Root_Rootf_Denisova.DG Denisova.DG edge 0.06184786910686927 NA NA
Root_Rootf_Denisova.DG Root_Rootf_Denisova.DGu edge 0.03156130870389199 NA NA
Roota X_pCE edge 6.018238691133988e-4 NA NA
Roota X_pCEl edge 0.0034956198406385143 NA NA
Root_Rootf_Denisova.DGu admixi admix 0.028839928870538642 0.028837540494757216 0.028839928870538642
Root_Rootf_Denisova.DGu Altai_Neanderthal.DG edge 0.025282785524437093 NA NA
X_pCE Czechia_ZlatyKun_IUP.AG.SG edge 0.13758138638526188 NA NA
X_pCE CE edge 0.0029056448407735057 NA NA
X_pCEl admixc_Georgia_UPb edge 0.006351128914891045 NA NA
X_pCEl admixn edge 0.007689299240307751 NA NA
admixc_Georgia_UPb admixc_Georgia_UPb_Iran_Wezmeh_N.SG edge 0.0028611346848520024 NA NA
admixc_Georgia_UPb admixfy admix 0.6969707635439528 0.6969520100236433 0.696980683259936
CE CEsb edge 0.0011059639271417886 NA NA
CE Russia_UstIshim_IUP.DG edge 0.0037836751861609606 NA NA
CEsb E3a edge 0.0047422501403539576 NA NA
CEsb admixi admix 0.9711600711294613 0.9711600711294613 0.9711624595052428
admixn Georgia_UP edge 0.11796551419563309 NA NA
admixn admixf admix 0.6042226001594933 0.6041878317024113 0.604231043220361
admixc_Georgia_UPb_Iran_Wezmeh_N.SG Iran_GanjDareh_N.AG edge 0.005397512527722755 NA NA
admixc_Georgia_UPb_Iran_Wezmeh_N.SG Iran_Wezmeh_N.SG edge 0.13060309079546814 NA NA
admixfy admixc_Georgia_UPbg edge 0.004267037956286423 NA NA
E3a E3ad edge 0.0044122247209337564 NA NA
E3a admixo edge 0.0027376829796744748 NA NA
admixi admixig edge 2.609153438398409e-4 NA NA
admixf X_pCE_Georgia_UPct edge 0.005451580143767893 NA NA
admixc_Georgia_UPbg admix admix 0.4599013851057849 0.45989109097180164 0.45990801771841056
admixc_Georgia_UPbg Iran_ShahrISokhta_BA2.AG edge 0.012161750571579075 NA NA
admix ILA.SG edge 0.01048434451163902 NA NA
admixh Russia_Yana_UP.SG edge 0.011515997952608449 NA NA
X_pCE_Georgia_UPct admixh admix 0.781453656372124 0.7814432483881505 0.7814729215795628
X_pCE_Georgia_UPct Russia_Sunghir_UP.SG edge 0.01419735946276054 NA NA
E3ad admix admix 0.5400986148942151 0.5400919822815895 0.5401089090281983
E3ad Onge_Jarawa edge 0.02244545343384189 NA NA
admixo E1_E2 edge 0.003928845578931799 NA NA
admixo E2_Onge_Jarawacy edge 0.003936513854457202 NA NA
E1_E2 Japan_Honshu_EarlyJomon.SG edge 0.032122427767155924 NA NA
E1_E2 E1_E2j_i edge 0.003071674141367137 NA NA
E2_Onge_Jarawacy admixfy admix 0.3030292364560472 0.303019316740064 0.30304798997635674
E2_Onge_Jarawacy E1_E2j edge 7.775220847112933e-4 NA NA
E1_E2j_i admixw admix 0.7331874782329287 0.7331166081095478 0.7333898926889186
E1_E2j_i Russia_DevilsCave_N.SG edge 0.02068681008414567 NA NA
E1_E2j E2_Onge_Jarawac edge 0.018411738075985687 NA NA
E1_E2j E1al edge 0.008279076591410792 NA NA
E2_Onge_Jarawac admixw admix 0.26681252176707126 0.26661010731108137 0.2668833918904522
E2_Onge_Jarawac China_Guangxi_Longlin_Epipaleolithic.AG edge 0.10529273873327694 NA NA
E1al admixh admix 0.21854634362787595 0.21852707842043717 0.2185567516118495
E1al admixhi admix 0.662571663640469 0.6625405057147175 0.662588920093644
admixw China_SEastAsia_Island_EN.AG edge 0.03175731252500638 NA NA
admixig admixigl edge 7.006299304370896e-4 NA NA
admixigl admixhi admix 0.33742833635953096 0.337411079906356 0.33745949428528255
admixhi China_UP edge 0.016210146827896028 NA NA
admixigl admixf admix 0.39577739984050675 0.39576895677963897 0.3958121682975887
admixig Bulgaria_BachoKiro_IUP.AG.BY.AA edge 0.07572554499987746 NA NA
Hoabinhian would be on the Onge_Jarawa branch but it had weird affinities due to low coverage or contam. Irula has an obvious preference for that branch but ShahrISokhta doesn't. I'm not sure if the extra admixture into AASI proposed earlier can even be detected by qpGraph. If the admixing population has no extra affinity to any samples, how would it show? Would it be obvious overfit in the residuals, or a huge branch length? And if it's most related to Iran then a double admixture edge from Iran is definitely unidentifiable, unless the Iran samples have different affinity to it.
Also tried adding Zongri to tease out "Tibetan ghost" if it exists but couldn't. Maybe because it's mostly YR derived. Both would be around DevilsCave/SEast_Asia territory.
Do we have actual evidence of admixture between humans and Neanderthals? Or between humans and Denisovans?
So far, all such evidence is quite weak.
It is true that Papuans have a higher percentage of Denisovan genes, but that does not mean that this is the result of direct admixture. Papuans separated from the rest of the tree and remained isolated for tens of thousands of years.
The percentage of archaic genes is very stable and it is the same in all human populations.
The same percentage is observed in geographically separated groups, the same percentage does not change much over time, starting from 50,000 years ago until now.
So why is it claimed that there was such admixture of humans with hominids?
I have done a lot of research in this area and my latest conclusion is that there is no such admixture. There is a deviation within a few percent, but it is not due to direct mixing, but is the result of quantitative accumulations and isolation, random processes in the distribution of a much more ancient initial population from which all humans originate.
(04-05-2025, 01:24 AM)TanTin Wrote: Do we have actual evidence of admixture between humans and Neanderthals? Or between humans and Denisovans?
So far, all such evidence is quite weak.
It is true that Papuans have a higher percentage of Denisovan genes, but that does not mean that this is the result of direct admixture. Papuans separated from the rest of the tree and remained isolated for tens of thousands of years.
The percentage of archaic genes is very stable and it is the same in all human populations.
The same percentage is observed in geographically separated groups, the same percentage does not change much over time, starting from 50,000 years ago until now.
So why is it claimed that there was such admixture of humans with hominids?
I have done a lot of research in this area and my latest conclusion is that there is no such admixture. There is a deviation within a few percent, but it is not due to direct mixing, but is the result of quantitative accumulations and isolation, random processes in the distribution of a much more ancient initial population from which all humans originate.
Neutral alleles are not under selection so the chance that drift alone is able to explain affinity is very dubious and unlikely. It's a different story for alleles under selection and those with LD, but most likely the alleles we do have nowadays are no longer under LD at least. Selection perhaps, but not most of them. Also it makes the most sense since the populations closest to Eurasians, Omotic people close to Mota, do not have Neanderthal affinity. Also Papuans are a pretty good piece of evidence since they definitely could not have gained more affinity toward Denisovans, considering how it is restricted to them and how significant it is. We also have some ancient samples with elevated Neanderthal and as expected the segments are longer than those in modern populations, which is indicative of admixture being closer in time.
