Can you define “Priors” and “Posterior inference” in a bit more detail, in the context of making a PRS for a specificic phenotype example?
Replying to “Can you define “Priors” and “Posterior inference” …”:
The priors here relate to the expected effect size - for example 90% of the snps might be expected to have no significant effect
Replying to “Can you define “Priors” and “Posterior inference” …”:
we assume a distribution for the SNP effects, i.e. each SNP effect is drawn from a distribution with mean mu and variance sigma. You can assume multiple distibutions, like x% is drawn from N(0,sigma1), y% from N(0,sigma2), etc. Then you update those distributions based on the observed estimated values in GWAS