search

bulik / ldsc

LD Score Regression (LDSC)
GNU General Public License v3.0
624 stars 339 forks source link

How can I perform genetic correlation of two traits from different population? European versus Asian. #52

Closed amarawi closed 7 years ago

rkwalters commented 8 years ago

Hi @amarawi, This kind of cross-population analysis is not currently possible with LD Score regression.

LD Score regression depends on the expected degree of tagging for polygenic signal based on the local LD structure, but that model of relationship with LD is violated when considering correlation for two traits with strongly differing LD structures. It's possible that the base LD score model can be adapted to correctly model this cross-ancestry structure, and that remains an active area of research, but for the moment there's no recommended way to implement this analysis in ldsc.

Cheers, Raymond

bulik commented 8 years ago

You might be interested in this pre-print

rkwalters commented 7 years ago

Hi, The pre-print Brendan mentioned is now published, and code is publicly available at: https://github.com/brielin/Popcorn

I'm closing this issue thread as inactive/resolved, but if you still have questions about this or any other issues feel free to follow up here or via the google group.

Cheers, Raymond