Do Northern Italians resemble more Germans, or Sicilians phenotypically?

[Grace O'Malley]

G25 is worthless because is undersampled, Bergamo is Lombardy, and as you can see the genetics of Italians is even more complicated.

That graph still shows the same thing. You can see that the only Italian region close to Germans is South Tirol which makes sense because they are Austrians or mostly descended from them. Population distances and similarity to populations is the same in G25 as other calculators like Gedmatch. Nothing really surprising there. Also some populations on the G25 have a very good amount of samples. There will also be more added and they are all academic samples.

[guezet]

That graph still shows the same thing. You can see that the only Italian region close to Germans is South Tirol which makes sense because they are Austrians or mostly descended from them. Population distances and similarity to populations is the same in G25 as other calculators like Gedmatch. Nothing really surprising there. Also some populations on the G25 have a very good amount of samples. There will also be more added and they are all academic samples.

Not exactly, the PCA shows more details that the G25 cannot show because it is undersampled. South Tirol is not close to the Germanic populations, it is just part of the Germanic populations. No surprises, they are ethnic Austrians.

[Vasconcelos]

G25 is worthless because is undersampled, Bergamo is Lombardy, and as you can see the genetics of Italians is even more complicated.

G25 was built with more references than that PCA itself, regardless of how many individuals you plot there. It's like building something with faulty tools. Besides there's only so much you can see in 2D, just because you plot closer to someone doesn't mean he's actually the closest sample to you.

[guezet]

G25 was built with more references than that PCA itself, regardless of how many individuals you plot there. It's like building something with faulty tools. Besides there's only so much you can see in 2D, just because you plot closer to someone doesn't mean he's actually the closest sample to you.

I guess the faulty tools are just the ones that show something you don't like.

[Vasconcelos]

I guess the faulty tools are just the ones that show something you don't like.

You're projecting, I dont give two shits who Italians (or anyone else, really) plot closest to. The reality is that G25 is a much better PCA than others, which is why the creator only uses that one, rather than the older PCAs.

If you're so interested in knowing, run a G25 check fit of north Italians using all relevant samples and see who comes up first.

[guezet]

You're projecting, I dont give two shits who Italians (or anyone else, really) plot closest to. The reality is that G25 is a much better PCA than others, which is why the creator only uses that one, rather than the older PCAs.

If you're so interested in knowing, run a G25 check fit of north Italians using all relevant samples and see who comes up first.

This is just your opinion, since the G25 is an amateur tool that in a few years will be overtaken by other amateur tools.

Whether this specific PCA is accurate or not, PCAs are used in academic studies, the G25 is not.

[Vasconcelos]

This is just your opinion, since the G25 is an amateur tool that in a few years will be overtaken by other amateur tools.

Whether this specific PCA is accurate or not, PCAs are used in academic studies, the G25 is not.

Because academic PCAs have thousands of components which make it's use extremelly slow for common folk. G25 yields results similar to the ones used by academics, as you can see in its plots. It's an amateur tool, but it's good enough for us. Certainly better than the older ones built by the same person, which is what you were trying to argue in the first place. It's not perfect, but so far it's the best tool (PCA) we have. If you find it unreliable, then just stop using any of these tools altogether

[guezet]

Because academic PCAs have thousands of components which make it's use extremelly slow for common folk. G25 yields results similar to the ones used by academics, as you can see in its plots. It's an amateur tool, but it's good enough for us. Certainly better than the older ones built by the same person, which is what you were trying to argue in the first place. It's not perfect, but so far it's the best tool (PCA) we have. If you find it unreliable, then just stop using any of these tools altogether

My point was completely different. You're defending the G25 for your own reasons that I don't even care about knowing.

[Vasconcelos]

My point was completely different. You're defending the G25 for your own reasons that I don't even care about knowing.

You have no point. It was built with more and better references than other PCAs such as the K13, K15 or K36, so it's a better tool for us amateurs-at-best to use, it's simple. If you don't like it, boohoo.

[Grace O'Malley]

Not exactly, the PCA shows more details that the G25 cannot show because it is undersampled. South Tirol is not close to the Germanic populations, it is just part of the Germanic populations. No surprises, they are ethnic Austrians.

That's just what I said. South Tirol are Austrians or Austrian descended.

G25 has a large German sample. It also has an East German sample. Italy is split into different areas because they have such differences between different parts of Italy. The benefit of something like G25 to people like myself is that anyone can use it. It is good at showing which populations are closest to each other. It's really brilliant as people can compare populations and themselves to modern populations and ancient samples. Most people now use the G25.