I have RNAseq HTSEQ count data for 3 individuals collected at 3 time points. I would like to perform count normalization across all 3 time points for each individual separately using Galaxy DESEQ2. So far I find that DESEQ2 is providing normalized counts for some samples but it outputs the original raw HTSEQ counts for others (2 out of 9 samples consistently have this issue). Has anybody encountered this before? Thanks, Sergey
Hello,
For your case, the experimental design may need some parameter tuning or a parameter intended to be set was missed. Double check how you set this up against the tutorials linked below to determine why the results differ between samples. I would also suggest reviewing the upstream jobs (mapper - HISAT2? and htseq_count itself) to ensure all data was processed the same way.
Galaxy tutorials for RNA-seq can be found here along with example data: https://galaxyproject.org/learn/
Those in the first group have been modified to work with the tools available at https://usegalaxy.org, for example: https://galaxyproject.org/tutorials/rb_rnaseq/
And this tutorial from the GTN, also linked from the page above, contains the same basic data from the usegalaxy.org tutorial, and also explains normalization methods in detail: https://galaxyproject.github.io/training-material/topics/usegalaxy/tutorials/rb-rnaseq/tutorial.html
Thanks! Jen, Galaxy team
Hi there, I am dealing with the same issue with my data, one of my replicates is not normalizing in DEseq2 even though all replicates were analyzed with the same settings. I was curious if you were about to resolve this issue or could provide more information. What I've tried so far: -Remapping the original .fasta file -Doing htseq-count again from the new alignment (counts matched the first attempt) -Using DEseq2 with all replicates plus this new one; found that this replicate was still the only one not normalizing.