what should be our references when 2 samples which I compare are treated and there is not any control?
1
0
Entering edit mode
@4fedfa78
Last seen 7 days ago
Japan

Hello everyone, I would like to use DESeq2 for my analysis but 2 samples are treated, imean there are 2 genes which I want to see the effects of drugs on both. but based on DESeq2, I must specified one as references as I am writing the codes as below, there is not any "control", so what should I do in this case?

Code should be placed in three backticks as shown below

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

# include your problematic code here with any corresponding output
# please also include the results of running the following in an R session

sessionInfo( )

DESeq2 • 196 views
0
Entering edit mode
swbarnes2 ★ 1.1k
@swbarnes2-14086
Last seen 12 hours ago
San Diego

You only have 2 samples? Total?

0
Entering edit mode

Actually, not in total, there are untreated samples too. but beside comparing treated and untreated I must compare 2 treated, to see drug affects on 2 different genes, it is my first time that I am doing compare 2 treated, what is the solution?

0
Entering edit mode

Read over the vignette. You can use contrast in the results() function. It is described in the vignette.

http://bioconductor.org/packages/devel/bioc/vignettes/DESeq2/inst/doc/DESeq2.html#contrasts

0
Entering edit mode

Many thanks for your always cooperation. I checked the link which you attached but I have a question, it means first I must follow the DESeq2 as usual and then after running at the result part I should use the "contrast" command as below?

results(dds, contrast=c("condition","C","B"))

and my another question is that because the condition is treated for both samples at all, so when I want to make my table.csv so should I name both treated for condition and what it means by "C","B" in the above command?

1
Entering edit mode

Hmm, if this section of the vignette is confusing, I think you'd be better off discussing with a local collaborator. It's important that you have a good understanding of what it means that you are forming a contrast of these two terms in a linear model.