7
First of all, I'm sorry for your loss and you can see this thread whenever you want, but I'm going to keep my promise that I'd post your results today.
Second (And anyone who is reading this):
The point of making these models is to try to recreate the qpAdm admixture methodology and try to lower the chances that a model could be rejected if it were made there.
Obviously we don't know if it can be rejected or not in G25, but that's why I try to reduce that possibility as much as possible following this points:
-Logical modelling.
-Use the same individuals and remove outliers as the studies/papers have stated.
-Not very ancient models.
-No more than 5 populations (Max. 5-way).
-Completely based on studies/papers of admixture models (But adapted to the person's roots).
-Play it safe.
If a model doesn't match with all those 6 requirements, I'd say that the admixture model posted here would be just for fun, so in consequence should not be taken seriously Imo.
Note: The results could change in case there are (Or find) new samples/coordinates that are not available for now in the G25 but are published in the genetic studies.
So taking all this text into account, I'm going to post your results in the next post.
Bookmarks