The error in G25 regarding Germanic admixture is that Etruscan_IA includes outliers which inflated the Germanic admixture. Modern Tuscans get only 9% Germanic and they more northern shifted than Medieval Tuscans.Etruscan,0.123751,0.1585354,0.0362664,-0.015818,0.0495476,-0.0101486,-0.0010771,-0.0037307,0.0235146,0.0446782,0.0005098,0.0109236,-0.0207298,-0.0060898,-0.0027332,-0.0033368,0.0030024,0.0005173,0.001222,-0.0043528,0.0011542,0.0045854,-0.0025266,-0.0042675,0.0010212
I don't have Medieval Tuscans but it should be closer to 5% than to 10% with this model which fits with Germanic Y-dna being more male-biased and pushing nearly 10%.
Target: Italian_Tuscany
Distance: 1.1331% / 0.0113305661.4 Etruscan 17.8 ARM_LBA 12.2 Levant_Ashkelon_IA2 8.6 Swedish
Also the percentages given for East Med admixture in the study for Imperial Tuscans (using ARM_LBA and other samples) are fine around 35-39% but one outlier pushed nearly 60% which screwed the whole result. Same with Germanic admixture, one outlier was more northern shifted, so scientists probably got their "20% Germanic" wrong from it.
If you check the PCA 3 out of 5 Imperial samples are very close probably those are around 39% East Med admixture.
[/COLOR]
NOTE: I am just showing a more proper way how could the Germanic admixture be estimated in the official studies model. This model is not necessarily meant to be taken literally.