I have data ( chromatin and gene expression ) from different tissues, in-house generated and from epigenome roadmap. I certainly have heterogenous samples with in same tissue. I want to normalise all the data so that homogenise the data from different tissues maintaining the variability across tissues. If I apply global normalisation methods, the biological variation across different tissues will be compromised. I am looking a way to normalised data with in each tissue. I found a program qsmooth that does kind of quantile normalisation with in each group, so normalises with in groups preserving the variation across groups.
In summary, does anyone have experience in analysing heterogenous data from multiple tissues ? Did anyone try qsmooth method ? I don't see many citations.