Is Jordan_ppnb better to use then natufian as an outgroup

[Genetics189291]

Which is better to use as an outgroup? Also if I have chg in my outgroup im guessing I don’t need Iran n.

[Kale]

CHG is quite distinct from Iran_N. Good to have both.
As far as Jordan_PPNB v. Natufian, neither are ideal, but they each have their redeeming qualities relative to each other.
Jordan_PPNB = less atrocious quality
Natufian = a more unique and drifted population

[Genetics189291]

CHG is quite distinct from Iran_N. Good to have both.
As far as Jordan_PPNB v. Natufian, neither are ideal, but they each have their redeeming qualities relative to each other.
Jordan_PPNB = less atrocious quality
Natufian = a more unique and drifted population

Of I had something like Jordan_ppnb, natufian and iraq_ppna would that be fine

[Genetics189291]

CHG is quite distinct from Iran_N. Good to have both.
As far as Jordan_PPNB v. Natufian, neither are ideal, but they each have their redeeming qualities relative to each other.
Jordan_PPNB = less atrocious quality
Natufian = a more unique and drifted population

This is my right populations now 

I’ll add Iran n but I can’t seem to get anything to pass I tried all combinations

[Kale]

If you update to admixtools 2.0.8, add the flag return_F4=TRUE, then you can see residual stats to see why a model is failing.

[Genetics189291]

If you update to admixtools 2.0.8, add the flag return_F4=TRUE, then you can see residual stats to see why a model is failing.

I managed to get two models to pass and it's always the italy early medieval samples for some reason, but I tried these same models with other North African groups which failed

Not sure why it's the only sample that passes for him, I tried arab and levantine combinations which all failed

[Genetics189291]

If you update to admixtools 2.0.8, add the flag return_F4=TRUE, then you can see residual stats to see why a model is failing.

This also seems to pass for my son, but it fails for the other two mozabite and saharawi

[Genetics189291]

I was having an issue with the standard errs relative to the weights so I swapped one of my right populations which iran_ganj_n.ag to iran_wezmeh_N.sg which seemed to sort out the issue standard errors and weights look good. I'll do a run on admixtools1 just to confirm