sgravel
commented
8 years ago I have never run more than 3 populations, because I have never found a situation where we could reliably assign more than 3 ancestries. I forget if the underlying code can handle more than three populations, I suspect that there might be a few hiccups if you did that, but they would be easily fixable. If you feel confident that you can reliably assign four local ancestries, I can help you make it work.
One workaround would be to combine two of the four ancestries (typically, the most difficult to distinguish) and treat them as a single ancestry.
On Sun, Nov 13, 2016 at 12:50 AM, wjaratlerdsiri notifications@github.com wrote:
Dear Simon,
I have tried to run my admixed population with 2 and 3-population models in your document. However, structure analyses indicated 4 ancestral populations in my cohort.
Could you suggest me how I can find a sample script for 4-population tracts analyses?
Thankyou for your great software and help.
Cheers, James
— You are receiving this because you are subscribed to this thread. Reply to this email directly, view it on GitHub https://github.com/sgravel/tracts/issues/8, or mute the thread https://github.com/notifications/unsubscribe-auth/AGnPhFQD7gb5z7xI3gWKAQbYVh3fU_lAks5q9qVCgaJpZM4Kwnct .
wjaratlerdsiri
Dear Simon,
I have tried to run my admixed population with 2 and 3-population models in your document. However, structure analyses indicated 4 ancestral populations in my cohort.
Could you suggest me how I can find a sample script for 4-population tracts analyses?
Thankyou for your great software and help.
Cheers, James