Averaging qPCR/Technical Repeats
Technical repeats such as qPCR repeats or RT repeats should be identified by a classification column in the data sheet, where the same number in the classification column indicates that the samples belong together. The figure below shows a part of a data file with qPCR replicates (#qPCR repeats).
Select Average technical repeats or Average qPCR repeats (the result is the same either way) in the Pre-processing menu in the Data editor to open a dialog box where the appropriate classification column is selected. The data is normalized by taking the arithmetic mean of all replicates. Once the data is normalized, the classification column is removed and the uppermost row label (sample name) in each group of replicates is kept. If technical repeats have been performed at several levels, e.g. both qPCR and RT, normalize the repeats sequentially starting at the lowest level. I.e. normalize qPCR repeats first, then RT, then sampling, then subjects etc.
When results will be compared in copy numbers (linear scale) instead of fold changes (logarithmic scale), expression of target genes should be normalized with reference genes before averaging RT (and earlier) repeats. The reason is that while qPCR repeats of target and reference genes are independent, RT repeats are paired (all target genes and all reference genes are measured for each RT repeat). Hence, in a study that has both RT and qPCR repeats and that will be analyzed in copy numbers (linear scale), the correct pre-processing scheme is to first average qPCR repeats, then normalize with reference genes, and finally average RT repeats. This reordering is not necessary when comparing data in logarithmic scale.