I want to estimate among-line variance in RNA-Seq gene expression between a panel of lines. The voom function outputs the expression values and their corresponding precision weights. I want to then scale the expression ranges of each gene so that they have a mean of zero and unit variance before fitting a mixed-effect linear model to each gene (to determine among-line variance). What does one do with the precision weights from the voom function when inputting it into the linear model? Does one keep the weights as is? Does one scale the precision weights for each observation?
If anyone could help please let me know :)