I have a data set of n=3 treated vs control RNA-seq. I used hisat2 for alignment followed by stringtie to compute fpkms, and then prepDE.py to convert the fpkms to count data, for analysis for D.E. using DESEQ2.
When I plot dendograms and heatmaps using hclust(ward), the results are different depending upon whether the input is rlog counts or log2fpkm.
I was wondering if anyone could help me interpret this- especially since 1 dendogram reflects the control vs treated grouping better. Thank you.