5
Thumbs Up |
Received: 2,840 Given: 842 |
Thumbs Up |
Received: 1,920 Given: 1,564 |
OK, princess, consider yourself special for making me repeat myself.
1) Use geographical/historical relevant populations for your own ancestral>modern ethnic group. There's no one-size-fits-all solution, West Asians should use Armenia_BA proxies, Eastern Euros should opt for adjacent Corded-Ware references, Western Euros for CW and Beakers, among others, etc.
2) Admixtools2 tutorial for windows.
Thumbs Up |
Received: 10,938 Given: 26,464 |
I've been following him for over a decade , for some reason he was obsessed with PCA plots since the very beginning (trying to prove that south europe is a levant-shifted version of N.Europe, using various pca plots, clines or whatever he calls those). I never understood that.
Thumbs Up |
Received: 528 Given: 787 |
Thumbs Up |
Received: 1,920 Given: 1,564 |
Thumbs Up |
Received: 10,938 Given: 26,464 |
Thumbs Up |
Received: 528 Given: 787 |
Thumbs Up |
Received: 1,920 Given: 1,564 |
I am not gonna do that.
The fact there's not a 'right-list'/outgroup, outside of admixtools usage, to check for validity means the results will always be skewed/overfit. But, don't get me wrong, PCAs like g25 have their usefulness, when you want a 'quick-and-dirty' method to get a rough idea of what you're dealing with in the first place.
Also, for moderns, PCA like g25, with good coverage to the template (davidski uses 23ame V3 raw template) is quite good, the problems pop-up with the ancient samples.
Thumbs Up |
Received: 1,806 Given: 953 |
Thumbs Up |
Received: 10,938 Given: 26,464 |
There are currently 1 users browsing this thread. (0 members and 1 guests)
Bookmarks