Skip to contents

Evaluates the coefficient from the linear regression of s2.post ~ sigmaSq. When there is no shrinkage, this value is 1. Values less than 1 indicate the amount of shrinkage.

Usage

shrinkageMetric(sigmaSq, s2.post)

Arguments

sigmaSq

maximum likelihood residual variance for every gene

s2.post

empirical Bayes posterior estimate of residual variance for every gene

Details

Shrinkage metric for eBayes quantifying the amount of shrinkage that is applied to shrink the maximum likelihood residual variance to the empirical Bayes posterior estimate