Using DESeq2 for DEG.
1
0
Entering edit mode
Nithisha ▴ 10
@nithisha-14272
Last seen 6.1 years ago

Hello all,

I have a dataset for which I have 6 samples- 2 controls, 2 for Treatment A and 2 for Treatment B. If I have a column in my Metadata called Treatment and it looked like this.  (The brackets are included in the name as well)

Treatment
Control (1)
Control (2)
Treatment A (1)
Treatment A(2)
Treatment B (1)
Treatment B (2)

And I wanted to compare control vs Treatment A and control vs Treatment B, would I be able to specify it under design= ~Treatment or would I have to use contrast? Or is it best to rename the rows altogether as Control, Treatment A and Treatment B?

Thanks!

 

deseq2 • 710 views
ADD COMMENT
0
Entering edit mode
@mikelove
Last seen 2 hours ago
United States

I'm concerned about providing DESeq2 code, knowing that this is a microarray dataset...

Help with carrying out DEG using Limma in R!

ADD COMMENT
0
Entering edit mode

Hi Michael,

Thank you for your concern. I have several datasets and the one concerning the question has raw counts and is not from the microarray experiment I had asked previously about! 

What I have tried so far is renaming the rows to Control, Control, Treatment A, Treatment A, Treatment B, Treatment B and then the following:

dds<- DESeqDataSetFromMatrix(countData=df,colData=mtfinal, design=~Treatment)

dds$Treatment <- factor(dds$Treatment, levels =c('control','Treatment A','Treatment B'))

dds<- DESeq(dds)

res1<-results(dds, contrast=c('Treatment','control','Treatment A'))

res2<-results(dds, contrast=c('Treatment','control','Treatment B'))

Is this the right way or is there a better method? Thank you for your help.

 

ADD REPLY
0
Entering edit mode

Ok, assuming you have RNA-seq counts, then for this experiment you would use a design of ~treatment, followed by DESeq() and results(). See the help page for ?results, in particular the 'contrast' argument as to how you could compare B vs control and A vs control. You can find all the information in the help pages:

?results
ADD REPLY
0
Entering edit mode

Thank you very much. If you do not mind, could you advice on how we would be able to tell if we were dealing with interaction terms? For instance, how can we tell if we should use 

design(dds) <- ~ genotype + condition + genotype:condition

or

dds$group <- factor(paste0(dds$genotype, dds$condition))

Should we already know for a fact that condition affects genotype to use the former?

Thanks.

ADD REPLY

Login before adding your answer.

Traffic: 782 users visited in the last hour
Help About
FAQ
Access RSS
API
Stats

Use of this site constitutes acceptance of our User Agreement and Privacy Policy.

Powered by the version 2.3.6