search

stephenslab / susieR

R package for "sum of single effects" regression.
https://stephenslab.github.io/susieR
Other
172 stars 43 forks source link

Does it matter to use R2 instead of R for LD in susie_rss #158

Closed XiaoweiHu-Stat closed 2 years ago

XiaoweiHu-Stat commented 2 years ago

I wonder if the results will be significantly affected by using squared correlation matrix R2 for LD in running susie_rss. Thank you.

gaow commented 2 years ago

@XiaoweiHu-Stat No, it is wrong to use R2. One must use R, the correlation.