Open garyzhubc opened 1 week ago
@garyzhubc I would not recommend using varbvs for identifying causal SNPs; for this, you should use susieR.
Hi Dr Carbonetto, Thanks for the feedback. Could you please explain why?
@garyzhubc Please see the susie paper or, for a shorter explanation, see Sec. 3.5.7 of the mr.ash paper.
Is it better to use PIP or Bayes factor for variable selection? I saw that the paper gives the option of computing Bayes factor of some null models. Would PIP or the Bayes factor give better result of finding which SNP is causal?