Riverman: I am all for freedom of speech, save the obvious (i.e. threatening speech) but your post about Rome being all over Europe, etc or genetically everyone is Roman kind of showed your hand, in my opinion. Your post started with this notion that West Sicily is fundamentally different from East Sicily, which is empirically nonsense. I am not going to do an entire literature review but going back to Sarno et al 2014, they found an I quote "our results point to a substantially homogeneous composition of maternal and paternal genetic pools within Sicily (East vs. West) as well as between Sicily and Southern Italy". Prior to 2014, there were the competing hypotheses that Phoenician presence on the West Coast vs. Greek colonization on the East, which never really got to where the City of Palermo is, might be reflected in differences of genetic pools. Well following Sarno et al 2014, Parolo et al (2016, Figure 1), Sazzini et al 2016 (Figure 2) and Raveane et al (2019), which has already been cited numerous times in this thread all find the same basic clustering of Italian populations across the 20 political regions, with Sicily clustering with Southern Regions. In addition, none of those papers found a difference in Sicily from West to East. So one can reasonably draw that you were perhaps trying to separate the Southern Italian regions from the ancient Romans. One could perhaps see that by first saying, well lets see if we can get West Sicily detached from East Sicily, then, that opens the door to say, well lets see if we can detach East Sicily from the Southern Mainland, and so on, and so on, till you get to a point hey Look, the Romans looked like us North of the Alps. You see we told ya so.
So I understand 1 paper publishes some findings, ok, but after 4 papers are finding the same results using different samples, different methodologies, then when I see an amateur calculator presented by someone on Youtube, or ItalianAnthro (the owner of that blog to his or her credit calls out the trolls quickly and to his/her credit, gives them a warning about presenting PCA plots from amateur calculators that are not consistent with the extant literature, if they persist after multiple warnings, the don't let the door hit you in the, you know the rest), etc, then I have to question the amateurs results. It could be an honest search, hence no integrity issues, or it could be someone with a political or troll agenda.
I made a statement earlier about the G25 model and after reading a post by Ygorcs I want to clarify. First, I have never ran G25 as I don't have my G25 coordinates, as I already stated. Second, if Ygorcs says the G25 model is a well specified statistical model, then I will accept what he said. I am not a trained statistician. I did take 21 hours of graduate research methods, statistics, Regression, Logistic Regression, 2-state least squares, etc, so I am a fairly good applied statistician (I am not a Statistician, to repeat). So if there are folks here who say the G25 Model is well specified, then I will take it as such.
However, If then G25 is in fact a well specified model, then when results are presented by amateurs using G25, or other such calculators, are not in line with a well established body of papers, then, I have to question the motives of such people. It could be, they are just honestly not aware of published papers, in that case, after being presented the evidence from the published papers, they revise their own G25 calculators before going to blogs, youtube, quora, etc and publish them, great, I respect that. However, if after being shown over, and over, and over, papers that are not in line with what there personal G25 models are saying, then I question their political motives (their Trolls). In the cases where the amateur G25 model is inconsistent with the published paper, and they know their results they produced are not in line with the extant literature, then it suggest to me someone data mining to find a desired result due to selecting samples that are from a extreme part of a distribution, e.g. from the extreme tail to outliers, or they are measuring variables inappropriately, etc.
As for Romans being genetically part of the DNA of occidental Europe. maybe, but did that shift them towards Italy, or more so Southern Italy even parts of Central Italy which cluster South? I don''t think I have drifted to far from Bell Beaker Sicily, or even Otzi the Iceman, which I share some pretty good amount of DNA with (per MTA Chroma analysis). Just take people like them add some Steppe and yes, maybe some Levant via Pheonicians, etc and that would be me. I think the 3 models from Dodecad12B, MDLP16 and Euro K13 pretty much show I am not too far from Iron Age Romans either, If I were 2 run Imperial Romans, get similar scores, run Medieval ROmans, similar scores. So you are going to have reconcile how Me, yeess Me (borrowing from Pink Floyd's The Wall yes Youuu), plot that close with those Roman samples.
So I am going to lay out my working hypothesis, and that's all it is, I think the Steppe Herders in Central Europe and more so in Nordic Europe got pushed to where they are due to the Steppers there having heavy, heavy EHG ancestry, not that me or other Sicilians or any other Italian gut shifted to where we are because of some massive migration into Sicily or Italy. So that is my story, and the first piece of evidence from Antonio et al 2019 seems to support my view, not that other research in the future with different samples could make it more nuanced, I understand that to be possible.