The following code in DE_timecourse (line 221) fixed the previous problem with voom weights, but didn't incorporate the user-defined values of Group and timepoint creating an error:
if(use_voom_weights){
design <- stats::model.matrix(
~Group*Timepoint + 0,
data=colData(object))
y <- edgeR::DGEList(counts=assay(object))
y <- edgeR::calcNormFactors(y, method="upperquartile")
v <- limma::voom(y, design, plot=FALSE)
weights <- v$weights
}
Need to have a test that runs voom weights and make tests run user-defined columns everywhere to trigger these problems in future.
The following code in
DE_timecourse
(line 221) fixed the previous problem with voom weights, but didn't incorporate the user-defined values of Group and timepoint creating an error:Need to have a test that runs voom weights and make tests run user-defined columns everywhere to trigger these problems in future.