Europe? This sounds like West Eurasia, not Europe. Natufian is a more basal older substrate that comes in by proxy through EEF and Steppe. Not a separate component in Europe.
If you're looking for a potential 4th, a more interesting example would be the ANE/Uralic/Nganasan pulse into northeast Europe that is a significant contributor to Finns and others in that region who are still very much European, but have a distinct signal that pulls them in an orthogonal direction on PCA plots.
This is just off the top of my head so I might be wrong, but I feel like there isn’t good justification for unpacking Natufian in this way by forcing a K=4 cluster analysis... like if we're unpacking, why not unpack Steppe into EHG and CHG, and EHG into WHG and ANE, with ANE also coming in via the earlier mentioned Nganasan/Uralic admixture into Finns, Estonians, etc.
I feel like the main 3 from Lazaridis 2014 are the best, and if we start unpacking we'll end up with 5 to 7, not those 4.
Hi Ansel. I didn't force 4 components. Admixture uses an unsupervised algorithm and 4 K provided the best fit. The sample is West Eurasian but the Natufian component is present also in Europe. It peaks in Neolithic Iraq. It popped up in different analyses. I will deal with these more in depth in Part II.
Thanks for the interesting and informative article.
Thank you! I will start working on Part II soon, where I will present some original results!
Europe? This sounds like West Eurasia, not Europe. Natufian is a more basal older substrate that comes in by proxy through EEF and Steppe. Not a separate component in Europe.
If you're looking for a potential 4th, a more interesting example would be the ANE/Uralic/Nganasan pulse into northeast Europe that is a significant contributor to Finns and others in that region who are still very much European, but have a distinct signal that pulls them in an orthogonal direction on PCA plots.
This is just off the top of my head so I might be wrong, but I feel like there isn’t good justification for unpacking Natufian in this way by forcing a K=4 cluster analysis... like if we're unpacking, why not unpack Steppe into EHG and CHG, and EHG into WHG and ANE, with ANE also coming in via the earlier mentioned Nganasan/Uralic admixture into Finns, Estonians, etc.
I feel like the main 3 from Lazaridis 2014 are the best, and if we start unpacking we'll end up with 5 to 7, not those 4.
Hi Ansel. I didn't force 4 components. Admixture uses an unsupervised algorithm and 4 K provided the best fit. The sample is West Eurasian but the Natufian component is present also in Europe. It peaks in Neolithic Iraq. It popped up in different analyses. I will deal with these more in depth in Part II.