When you perform batch correction using RUV normalization, what do you use as an input for the RUVphospho() function? Do you use a matrix of log-transformed raw peptide intensities or do you first normalize the log-transformed intensities using Z-normalization or median normalization?

Another naive question for a non-bioinformatician: What does the medianScaling() function exactly do? Does it subract the median of each sample from each data point in that particular sample and divide this by the sample's median absolute deviation? If you can provide a mathematical formula for the function, that would be very useful.

Thank you very much.