Search results

I'm definitely looking at this way too long and should take a break. But the Illyrian sample... I removed Franks from what may not have even been a good model to begin with (i.e. my modifications). Which broke down into Germanic (>Angles), Celt (>Pict), Iberian, and Illyrian? Which was...

Did so with Scandinavian, Langobard, and Goth removed. Enough overlap with Sca/Jute, and probably not too much Langobard or Goth in Britain. Basically Jute and Frank are perhaps proxies for Scandinavian and more continental Germanic. Seems like a pretty balanced mix. You might also see how...

I'm on board. South Gaul is just my big fingers ☺️ The ANE however is new. The good old days in Siberia, perhaps. Extremely remotely maybe Denisovan overlap, maybe a Paupan sample if it's not already there, would appear

Seems to play on more or less the theme: 2/3 Germanic, fairly Scandinavian but falls continental 1/3 Celt, pretty balanced but perhaps slightly centered on Britain/Wales Comparing the two coordinates these are more Germanic. But even here lumping samples is almost 50/50 Chebyshev with 3 pops...

I guess it's interesting that everyone is researching for a different reason. For me my questions are very small or match-based. Just for your own comparison with these new coordinates and reckless Chebyshev experiments it similarly likes to use especially South American to balance. First time...

It's all just for fun for me. At any rate DNA has lots of unfun applications. Do I grow toenails like a Celt or a Germanic? It's all merely to do with piss and blood. I imagine, whatever your side or vision, that the uses of DNA once it becomes powerful will sadly become commonplace, political...

This is only one example but a fit of 0.007 and the V1.1 model prefers Scandinavian over Germanic because obviously Jutes were north shifted, so Chebyshev here will do about 2/3 Sca 1/3 Ger instead of pure Angle/Jute No idea why it switches Briton for Wales Then the models can be cleaned up...

I'll hopefully leave the matter alone after this. But here are some Chebyshev experiments with fits of 0.002 sometimes. With simulated Coords there is now a mysterious Indus component. Likely noise or perhaps the trace Roma floating around. Also it interestingly/importantly swaps Goth for...

Two-way has a lot more depth, too. V2 w/o 'Scandinavian'

I'm East Coast and should probably sleep But one more for tonight is, without 'Scandinavians.' It seems to absorb a lot of the Germanic, perhaps rightly so. Anyway:

Here's those. It's amazing to get results that make sense with these simulated coordinates.

And the SSA quirk vanishes with these new coordinates. Who knows what's really correct but I have some serious doubts about IllustrativeDNA, at least in my instance. Or maybe simulated coordinates are over-fitting 🤷

Another comparison. One is IllustrativeDNA. The new, seemingly much more accurate/nuanced is ExploreYourDNA Edit

This is pretty remarkable. Same RAW file, two G25 compositions

Fairly Scandinavian weighted. Celt seems underrepresented in the spread in this specific instance Saxon closeness (between the three we are taught as a sequence, Angles, Saxons, Kites) is a common theme. Welsh underrepresentation is a theme too though I have 'recent' North Welsh via Hughes and...

On second look I think I'm misinterpreting the word "distant" as time, instead of genetically distant. I still value the tool and even the Chebyshev over-fits, partly because there seems something honest about them, in the sense that we are pieced together from all over. And how these pieces...

Agreed. Most of them leave me somewhat wondering... However there seems to be this very critical advantage: "Adjusting the distance parameter in the analysis will change the number of samples in the output. A smaller distance will result in more samples and more distant ancestry being included...

This evening's models were irresponsible, almost certainly misleading, incorrect, and poorly arranged. Happy to share them. Interestingly I read distance changes the "timeline", that 0.05 has an older perspective than 2.0. All along I thought it had to do with grouping of results. Or perhaps...

I do think that's spot on. Going by the lack of matches, and various data... *5.5% Baltic MyHeritage *2% Baltic FTDNA *2.4% Baltic Hunter Gatherer *Closest matches to VK Saaremaa samples I'd deduce it's almost none to do with modern times. And about half has to do with Viking age travels, half...

Then again modeled by Lithuania. It seems as Germanics they were closer to Danes than Swedes properly, and likewise had some similarity to Finns and Balts. But anyway that's my on going task. Find out about this very specific branch, seemingly, apparently, from a Bashkir and a person from...