Understanding how to contrast two treatments in DESeq2
2
0
Entering edit mode
Jennifer M • 0
@16fd24f8
Last seen 3 days ago
United States

Hi all, I want to confirm that I am interpreting my results correctly. I am using DESeq2 to find differentially expressed genes and using the following code to extract results from the comparison between two treatments:

res<-results(dds2, contrast = c("Condition", "Control", "Treatment"))

write.csv(as.data.frame(resOrdered), file="savedfile.csv"


If I have gene A that has a log2foldchange of -2.0, that can be interpreted as "gene A was downregulated in the treatment group compared to the control", yes? I am concerned that I am not making the correct interpretations based on the order of my groups when I run the code. I can swap the position of "Control" and "Treatment," and gene A would then have a log2foldchange of +2.0, so I have assumed that whatever group I have on the right-hand side is the causative of up or downregulation comparatively. Thank you!

DESeq2 RNASeqData • 154 views
1
Entering edit mode
ATpoint ★ 1.6k
@atpoint-13662
Last seen 15 hours ago
Germany

Your interpretation seems correct. You can always use plotCounts() or any custom plotting based on the normalized counts to confirm things like that.

0
Entering edit mode
swbarnes2 ★ 1.1k
@swbarnes2-14086
Last seen 8 hours ago
San Diego

Most people follow the convention of comparing to the control, so most people would do

res<-results(dds2, contrast = c("Condition", "Treatment", "Control"))


And say the Treated is upregulated compared to Control. But it's just a conventional preference.