I have an issue with the density plot of my microarray data. My data span 6 chips of the Illumina HT12 v4 beadarray. I uploaded all the raw data to Illumina IGS and extracted the SampleProbeProfile and ControlProbeProfile. I proceeded with Limma neqc normalization and transformation. The PCA plot puts the data in two clearly distinct clusters. The density plots of the normalized data are normal distributions that form two separate peaks as well: some lines have one pick and some another.These are corresponding to the two different clusters. But can I continue using these data or not? I tried standardizing them (mean=0, sd=1) and the two peaks remain present but this time they are much closer. Shall I use this second transformation?