hi
I am writing to you with a doubt/question regarding the heatmap visualization of gene expression data obtained with bulk RNA-seq technology.
In particular, my analysis aims to investigate the possible similarity in the expression profiles between my cellular model and other cells whose profiles are present in databases available online. To do it, I started from the fast files from my experiment and other datasets and performed the alignment and the calculation of the rlog normalized value uniformly for all the datasets used.
However, once I create the heatmap and scale the gene values via z-score, the heatmap shows the samples belonging to the same dataset as having the same expression profile, while the samples from different datasets seem to have different profiles. In particular, by using the same list of genes, the heatmap generated by using only samples from my experiment showed clear difference in the expression of these genes between patients vs controls
; but when I compared these expression levels with those of other cells and I create a new heatmap it seems that differences between samples and controls disappear, while there seem to be opposite differences in expression between samples from different datasets
(making me suspect that this is a bias related to normalization with the z score). can you give me some suggestions on how to solve this problem? Thanks