Dear @aljabadi,
yes I do have different plots from your code though my initial question was regarding the ‘non-changing’ Color key scale (being always between -2 and 2 regardless the ncomp and data plotted with cimDiablo). Furthermore, I was getting confused after reading the rCC greater than 1 discussion as I wasn’t sure if such Color Key Scale was indicating correlations or not (which it’s now clear it is not).
What I would still like to double check, if possible, it’s the correct interpretation of the cimDiablo results as for my previous question:
@joosefupas
what does the Color Key scale indicate? Some sort of differential expression in terms of fold change differences (starting from my previously log-transformed and normalised data in my case)
For example, taking your nutrimouse dummy example (hereby depicted in the screenshot), how should I interpret the red highlighted blocks (e.g. such selected features were found being ‘overexpressed’ [as their Color Key goes towards 2 i.e. red color in the Color Key Scale] respectively in coc, fish, and lin [rows categories for nutrimouse data] as shown after such features are hierarchically clustered, right?
Thanks a lot again
Joosè