Both "Genetiker" and "Felix" and perhaps "Eurogenes" have run the ancient genomes in question through various calculators and programs including Oracle to get a handle on the similarity of these genomes to modern populations.
Predictably enough, the percentages differ depending on who did the analysis. Instead of trying to come to some degree of consensus, they just seem to be engaging in a *******contest.
Using the percentages for the K23b and v3 (Dodecad K-12) runs provided by Genetiker (along with the results from the Dodecad K12b run) Maciamo analyzed the data and concluded that ERS389798, Hinxton 4, was more likely to be one of the Iron Age "Celts", and Hinxton 1 was more likely to be from the Anglo-Saxon period.
On the thread started by Fire Haired I took the attribution given in other places, looked at the same percentages that Maciamo had looked at, and concluded that those percentages supported an attribution of Hinxton 4 to the Iron Age Celtic period more so than did the data for Hinxton 1.
I said the following:
"The more northern, more northeastern "tilt" of the "Anglo-Saxon" sample seems pretty clear."
I also said that I would find it surprising that the AngloSaxons would turn out to be the more western and southern shifted group.
This was my reasoning based on those runs:
"The Anglo-Saxon period male has approximately the same amount of "Caucasus" as the Iron Age Celt, and only 2.5 points less Atlantic Med. However, he is 11 points more "North European". He also has 3.63% East Asian, and 1.16% Siberian, none of which show up in the Iron Age Celt.
The Iron Age Celt has 2.63% Northwest African, compared to 1.31%, he has 2.73% East African, compared to .79%, and he has 5.96% Gedrosia and 2.90% South Asian, compared to virtually none for the Anglo-Saxon.
So, once again, of the two, and taking into account the abstract of the paper, the Hinxton 4 sample seemed to be the likelier candidate for an Iron Age Celt.
Who knows, maybe they were both from the Iron Age. The authors will tell us. Maybe Hinxton 1 was of mixed ancestry. I don't know.
That's it...that's what drew the fire storm. Apparently, just using a Dodecad calculator is enough to draw the ire and insults of a certain group of posters, even to the point of calling this site a joke.
As I stated upthread, new percentages have now been provided by "Felix" for Hinxton 4. I don't have the data for Hinxton 1 so I can't do the same kind of comparison. What I can do, and did, is look at the percentages in the v3 spreadsheet. (As I said, I no longer have DIY or the Oracle set up on my computer.) Anyone can do the same by clicking on the link. So far as I can see, the ancient sample is pretty close to the western Scots (Argyle and Orkney), except that it is 2-3% more Med than the modern populations, and 2-3% less West Euro. Considering the documented population movements into those areas since the Iron Age, I don't think that's surprising.
Since then I've seen a post on another site to the effect that the sample is closest to the Irish. That may or may not be an accurate description of the results. I also don't know if that's based on an Oracle run or shared drift analysis. I have no way of knowing. This must be from a Eurogenes analysis, because in the Dodecad runs, the Irish do not seem to be as close to the ancient sample as the Scots, probably because they have less "Eastern European." (3% to 11%, with the sample having 11%) I don't know what the population averages for these modern groups are in the Eurogenes K-13 run which does have Western European and Eastern European clusters.
I also think that I saw a Eurogenes K13 analysis of some of these ancient samples, perhaps on Anthrogenica, but I don't remember precisely where. If I have time I will check to see how much Eastern Euro each one had.
At any rate, what I don't understand is why the abstract implied that the Anglo Saxon period people were more like modern Brits than the Iron Age people. If Hinxton 4 is indeed Iron Age, he looks pretty darn close to modern Brits, at least of the Scots variety.
Ed. Using the K23b percentages provided by "Felix", which are different than the ones provided by "Genetiker", and if you add what he calls EEF and Caucasus together, (Caucasus being, perhaps an eastern drifted from of EEF) you get an EEF/WHG/ANE set of numbers for Hinxton 4 which are only a few points off from the numbers in the Lazaridis et al paper. (For example, you get an EEF approximation of 47% compared to 49.5%) It would be nice if the authors used the Lazaridis algorithm on these samples so we can make a real comparison.