Unmasking AIT

[Husky]

Post 2/3

of archiving the parts that struck me as relevant in the actual study's paper.



Still taken from nature.com/ncomms/2015/150519/ncomms8152/full/ncomms8152.html

[...]

Considering haplogroups R1b-M269, R1a-M198 and I1-M253, and the 95% highest posterior density intervals of their TMRCAs, 64% of the MSY sequences sampled in our study descend from three ancestors who each lived more recently than ~7.3 KYA.



(The above statement is what all the news items on the subject talked about: 64% or ~2/3 of European males descend from 3 ancestors - distinguished with regard to haplogroups R1b-M269, R1a-M198 and I1-M253 - who came after 7300 years ago.)



[...]

Bayesian skyline plots (BSPs) (Fig. 2) reveal the variation of effective population size with time21. The plots are consistent with patterns seen in the relative numbers of singletons, described above, in that the Saami and Palestinians show markedly different demographic histories compared with the rest, featuring very recent reductions, while [size="5"]the Turks and Greeks show evidence of general expansion, with increased growth rate around 14 KYA. A different pattern is seen in the remaining majority (13/17) of populations, which share remarkably similar histories[/size] featuring a minimum effective population size ~2.1–4.2 KYA (considering the 95% confidence intervals (CIs) reported in Supplementary Table 4), followed by expansion to the present. Considering only these 13 populations, the only significant geographical correlation is of decreasing diversity in the number of polymorphic sites from east to west** (Supplementary Table 3); notably, there is no significant correlation between the age at which effective population size was at a minimum before expansion (Supplementary Table 4), and either latitude or longitude (Supplementary Table 3). Taken together, the very recent age of the demographic shift and its lack of geographical pattern suggest that its origin is distinct from that of the diffusion of agriculture.

IMO, the extremely-emphasised portion immediately above is possibly the most revealing statement in the entire article: that Greece is NOT included in the 13 populations who share a different history/were powerfully and often fundamentally influenced by the Bronze Age invasions.

That is, that the Bronze Age history of Greece is different from that of the other European "IE" places considered. Surely it means that the Greeks did not share much in the "Bronze Age Riders=Oryans from the Steppe=urheimat" history?



** "Considering only these 13 populations, the only significant geographical correlation is of decreasing diversity in the number of polymorphic sites from east to west"

Meaning... less genetic diversity seen in the west of Europe rather to the east? Which underlines that the Bronze Age spread effects discussed are from E to W.

The proportion of the parameter variance explained by the summary statistics (R2) is in most cases higher than 10% (and hence generally considered as a good estimation), but the 95% credible intervals are wide. This is particularly true for T1, the time of the start of the demographic change (reduction or expansion), thus preventing us from drawing any conclusion about the timing of these events (whether post Neolithic, or more ancient) from the ABC analysis.

In general, the non-parametric genealogical approach represented by BSPs better explores the variation found in our data, compared with a more conservative ABC analysis based on a single locus. We note that both analyses assume panmixia, and that population structure might influence effective population size estimates. However, it seems improbable that such an effect would extend to so many populations.

Panmixia - the willingness to mate with anyone - here may not be in conflict with the social stratification of Oryans (as per Oryanism).

Since the study looks at the male-specific region of the Y chromosome, i.e. concerns paternal lines, can assume there are fewer restrictions for "IE" males to mate with anyone. -> The "thundering down on chariots and stealing native females" scenario that Oryanists always fondly 'reminisce'=fantasise about.

Our approach has led to the confident identification of many MSY sequence variants in European population samples and a highly resolved phylogeny, but our conclusions are also influenced by a more contentious factor, the choice of mutation rate.** We chose a rate (1.0 (0.92–1.09) × 10-9 per bp per year, considering a 30-year generation time) based on the observation of 609 MSY mutations (excluding palindromic regions) in Icelandic deep-rooting pedigrees24. The point estimate of this rate is the same as an earlier pedigree-based estimate in which only four mutations were observed25, and which we applied in our broader MSY-phylogeny study13.



We note that the rate we have used is higher than the estimate of 0.76 (0.67–0.86) × 10-9 per bp per year based on counting the ‘missing’ mutations in the genome of the Ust’-Ishim male26, radiocarbon dated to ~45,000 YBP. Other studies27, 28 have inferred slower mutation rates (0.62 or 0.64 × 10-9) based on scaling the genome-wide de novo rate to account for male-specific transmission, though this has been criticized29, and is not consistent with phylogenetic mutation rates estimated from human–chimpanzee MSY comparisons30, 31. Some have chosen to calibrate the pedigree mutation rate against external (for example, archaeological) data15, 32, but we have rejected this idea, firstly because of uncertainty over how archaeological date estimates correlate with demographic changes, and secondly because we have used a coalescent-based dating method that itself models genealogies. Despite recent advances, mutation rate remains a difficult issue, and more data are needed.



(** Probably not offensive. But they themselves think it's a question mark. Perhaps something that will be refined in future studies, which may even redo the dates.)



[...]



The recent and rapid continent-wide demographic changes we observe suggest a remarkably widespread transition affecting paternal lineages. This picture is confirmed in an independent analysis of MSY diversity in the pooled HGDP CEPH panel European samples16, and is compatible with current (n=98) ancient DNA data for MSY (Fig. 3; Supplementary Table 8), in which hgs R1a, R1b and I1 are absent or rare in sites dating before 5 KYA, whereas hgs G2a and I2 are prevalent.



Figure 3: Timeline of MSY ancient DNA data.



The graph shows stacked frequencies of MSY haplogroups in ancient European DNA samples, based on data from 98 individuals, and binned into 500-year intervals. ‘other’ includes C1, F*, H2, R*, R1, R1b(xR1b1a). Below the timeline are indicated BEAST point estimates and highest posterior density intervals for three relevant haplogroups.

Very much worth looking at Figure 3 too at the link:

nature.com/ncomms/2015/150519/ncomms8152/fig_tab/ncomms8152_F3.html



Figure 3:



- doesn't show the aquamarine colour blue of R1b-M269 that is the most common blue=R1* in Greeks,



- and aquamarine blue R1b moreover shows what looks like the earliest divergence compared to the other R1b colours in Figure 1.



- Figure 3 only shows R1a(-M198) - the lightest blue - which is about 1/18th of the Greek pie-chart.



- And assuming I understand figure 3 correctly: as per Figure 3, the lightest blue=R1a(-M198) already existed in European MSY DNA at ~7,250 years ago. I.e. well before the steppe Bronze Age Riders' massive migration (replacement is invasion, IMO) into Europe.

So the least that can then be concluded is that R1a didn't enter Europe in the Bronze Age. I make no guesses as to whenever or wherever it entered from. But if this R1a existed at say 7300 years ago in Europe (since that's the earliest date given for the 3 paternal lineages that are to account for 2/3rds of modern European Y/males), this R1a wouldn't have first been brought to Europe in any European Bronze Age massive migration from the steppes. So only 2 out of the 3 paternal lineages of European Y can then have a putative European Bronze Age origin in the steppe.

So what then accounts for the earlier R1a spread in Europe? After all, everyone said that R1a (like R1b) was one of the invasionist gene sequences. And it is still implied, with R1a being one of the 3 paternal lineages associated with the Bronze Age "changes in burial" and "spread of horse riding" etc in Europe.

So Anatolia will then have to be argued as Urheimat with IE spread from diffusion from farming around 7000 BCE: a farming migration can't be from the steppe - Eurasian steppe nomads is what Oryans from a steppe urheimat were ever considered, dubbed pastoralists (though others domesticated the horse). The PIE-ist argument against Renfrew's Anatolian hypothesis was that his date for PIE was too early, that PIE should postdate farming.

But then: R1a within Europe itself predates Kurgan (steppe) PIE then... <-> How is R1a an invasion marker then?

Kurgan Urheimat is dated 5000 BCE. Mallory dates the "IE dispersals" into Europe starting at 4500 BCE, earlier than Gimbutas' ~3000 BCE. R1a in Europe at 7250 years BP = 5250 BCE is then older than the Steppe Urheimat/PIE at 5000 BCE.





Continued in next.