Hello,

I am new using DESeq2 on R. I always used galaxy. My question is... isn't the comparison supposed to be between 

~condition treated vs. untreated or knockdown vs wildtype?

When I use R and type :

> resultsNames(dds)
[1] "Intercept"                      "condition_untreated_vs_treated"

> resLFC
log2 fold change (MAP): condition untreated vs treated 
Wald test p-value: condition untreated vs treated 

even with the pasilla example, I get untreated vs treated when in the manual appears treated vs untreated. Am I doing something wrong?

One of the genes with the most significant change that I am looking at appears downregulated when I use galaxy, but upregulated when I use R.

Who should I trust?

Thank You!

It seems that you are having a problem with setting the reference level. You can make sure that the untreated level is the reference level by setting 

dds$condition <- relevel(dds$condition, ref = "untreated")

before running DESeq() as is mentioned in the vignette.

-R