search

stephenslab / susieR

R package for "sum of single effects" regression.
https://stephenslab.github.io/susieR
Other
174 stars 44 forks source link

Is Susie sensitive to selection bias? #99

Closed guillaume-ramstein closed 4 years ago

guillaume-ramstein commented 4 years ago

I would like to use SuSie to select candidates from a set of preselected markers (~100 preselected SNPs out of millions). One potential issue: markers were preselected for genome-wide significance (in GWAS). Can I still use SuSie, even though I'm choosing candidate causal variants from a nonrandom subset of markers?

stephens999 commented 4 years ago

no, you can't use SuSie here.

On Tue, Mar 10, 2020 at 2:54 PM guillaume-ramstein notifications@github.com wrote:

I would like to use SuSie to select candidates from a set of preselected markers (~100 preselected SNPs out of millions). One potential issue: markers were preselected for genome-wide significance (in GWAS). Can I still use SuSie, even though I'm choosing candidate causal variants from a nonrandom subset of markers?

— You are receiving this because you are subscribed to this thread. Reply to this email directly, view it on GitHub https://github.com/stephenslab/susieR/issues/99?email_source=notifications&email_token=AANXRRIKPW3FHCJJASOLIQDRG2LIVA5CNFSM4LFGX6S2YY3PNVWWK3TUL52HS4DFUVEXG43VMWVGG33NNVSW45C7NFSM4IUAE2YQ, or unsubscribe https://github.com/notifications/unsubscribe-auth/AANXRRL5TGYCTOOF26BQW6LRG2LIVANCNFSM4LFGX6SQ .