Hi all! I have a set of data, and I have normalized them using a couple of normalizations. I modelled MSE and calculated F-ratio (signal-to-noise ratio) for each normalization method. Should I go for normalizations that lead to lower MSE or calculated F-ratio? (MSE is modelled just between replicateexperiments like:A" Comparison of Normalization Techniques for MicroRNA Microarray Data", F-ratio is calculated with a formula that compares MSQ between and within conditions, as in Comparison of normalization methods for Illumina BeadChip HumanHT-12 v3)