What's the concensus on unscaled coordinates?

[Dukaginzade]

For me, they work much better.

My fits on scaled are absolutely terrible, but when I use other people's g25 unscaled coordinates, it completely messes up and assigns random ethnic groups.

I myself am not certain about some parts of my lineage, so I was wondering how trustworthy unscaled coordinates are? I feel like it helps me, but everyone I used can be modelled with the ancestry that I have weirdly enough, which makes me doubt my research.

I use them for modern admixture. Not for ancient.

[celticdragongod]

For me, they work much better, my fits on scaled are absolutely terrible, but when I use other people's g25 unscaled coordinates, it completely messes up and assigns random ethnic groups.

I myself am not certain about some parts of my lineage, so I was wondering how trustworthy unscaled coordinates are? I feel like it helps me, but everyone I used can be modelled with the ancestry that I have weirdly enough, which makes me doubt my research.

I use them for modern admixture. Not for ancient.

I think most people here see both scaled and unscaled as largely for entertainment purposes. You really shouldn't take either too seriously.

[Dukaginzade]

Okay, so what's the most reliable way to figure out if a certain admixture is correct?

[ScandinavianCelt]

For me, they work much better.

My fits on scaled are absolutely terrible, but when I use other people's g25 unscaled coordinates, it completely messes up and assigns random ethnic groups.

I myself am not certain about some parts of my lineage, so I was wondering how trustworthy unscaled coordinates are? I feel like it helps me, but everyone I used can be modelled with the ancestry that I have weirdly enough, which makes me doubt my research.

I use them for modern admixture. Not for ancient.

Do you have a unscaled coordinates for each chromosome? I find that they help to refine your understanding of your inheritance by chromosome more than a composite unscaled.

[djipon]

Do you have a unscaled coordinates for each chromosome? I find that they help to refine your understanding of your inheritance by chromosome more than a composite unscaled.

Where can you get such coordinates? Like real ones... I have the ones from DNAgenics, but I know they are simulated.

[ScandinavianCelt]

Where can you get such coordinates? Like real ones... I have the ones from DNAgenics, but I know they are simulated.

I wouldn't worry about DNAGenics like so many others do unless you have some glaring difference in your composite PCA plots. Mine are virtually identical and therefore I trust my 22-chromosome coordinates with DNAGenics should look the same as IllustrativeDNA if they offered us these-- and they don't, apparently.

Some people's plots might not look the same as mine in terms of closeness-- maybe it depends upon the raw data they have to work with?

How do your composite UNscaled compare on the PCA? Mine look like this:

screenshot software for windows

[djipon]

I wouldn't worry about DNAGenics like so many others do unless you have some glaring difference in your composite PCA plots. Mine are virtually identical and therefore I trust my 22-chromosome coordinates with DNAGenics should look the same as IllustrativeDNA if they offered us these-- and they don't, apparently.

Some people's plots might not look the same as mine in terms of closeness-- maybe it depends upon the raw data they have to work with?

How do your composite UNscaled compare on the PCA? Mine look like this:

screenshot software for windows

There you go

[Dukaginzade]

No I don't. How can I break them down one by one?

[JerryS.]

Okay, so what's the most reliable way to figure out if a certain admixture is correct?

I'd like to know this too...

[Dukaginzade]

Bump