I was testing the algorithm on the provided sample data, and noticed that modification of prior.1 from 1e-04 to 1e-05 increases the posterior probabilities, which is not the direction I would expect. Note, this is using the default of variant specific priors rather than uniform priors. A reproducible example here:
daghli
commented
5 years ago
Hi,
I was testing the algorithm on the provided sample data, and noticed that modification of prior.1 from 1e-04 to 1e-05 increases the posterior probabilities, which is not the direction I would expect. Note, this is using the default of variant specific priors rather than uniform priors. A reproducible example here:
library(hyprcoloc) betas <- hyprcoloc::test.betas ses <- hyprcoloc::test.ses traits <- paste0("T", 1:10) rsid <- rownames(betas) results<-hyprcoloc(betas, ses, trait.names=traits, snp.id=rsid,prior.1=1-04);results results<-hyprcoloc(betas, ses, trait.names=traits, snp.id=rsid,prior.1=1-05);results
Is there any explanation for this behavior? Are my prior (no pun intended) conceptions incorrect?
Thank you.