search

stephenslab / susieR

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

Reporting per-snp BFs #109

Closed trb8 closed 3 years ago

trb8 commented 4 years ago

Hi Thanks for the great method/package. Will it be possible to report the per-SNP BFs (for each single effect) using e.g. s$snplbf[l,] <- res$lbf in update_each_effect_ss

Those BF will be useful for coloc analyses as in COLOC here . The COLOC method itself assumes a single effect so there is a need.

Thanks, Tushar

stephens999 commented 4 years ago

I agree this would be useful, and is certainly something we have used ourselves in the past. If it is not yet an option we should add it.

On Fri, Aug 28, 2020 at 1:01 PM trb8 notifications@github.com wrote:

Hi Thanks for the great method/package. Will it be possible to report the per-SNP BFs (for each single effect) using e.g. s$snplbf[l,] <- res$lbf in update_each_effect_ss

Those BF will be useful for coloc analyses as in COLOC here https://journals.plos.org/plosgenetics/article/file?id=10.1371/journal.pgen.1004383.s015&type=supplementary . The COLOC method itself assumes a single effect so there is a need.

Thanks, Tushar

— 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/109, or unsubscribe https://github.com/notifications/unsubscribe-auth/AANXRRPLBIVUHDNQ242EOXDSC7WJHANCNFSM4QOMB3OA .

trb8 commented 4 years ago

Great! Thank you.

pcarbo commented 3 years ago

Hi @trb8. I just want to make sure I understand your request.

The idea would be to create an output, let's call it lbf, that is a p x L matrix, where p is the number of variables and L is the number of single effects, in which lbf[i,j] is the log-Bayes factor for variable i and single effect j. Is that what you looking for?

gaow commented 3 years ago

@pcarbo sorry I have added it earlier but not tests and documentations to it yet ... I'll handle this and other issues #111 this evening on my calendar.

trb8 commented 3 years ago

@pcarbo yes, that's what I meant Thanks @gaow

pcarbo commented 3 years ago

@trb8 susie and susie_rss now include an output lbf_variable. It is a matrix containing the per-SNP Bayes factors. Please let us know if you have any questions about this, or anything else.

pcarbo commented 3 years ago

@trb8 I neglected to mention that this is implemented in the cran branch.