Hi all,

I have affected vs control samples collected at three different time points, and I would like to compare differentially expressed miRNAs between affected and control at different time points (t1, t2, t3). Here is my sampleTable:

> sampleTable
      condition timepoints
A1_T1  affected         t1
A1_T2  affected         t2
A1_T3  affected         t3
A2_T1  affected         t1
A2_T2  affected         t2
A3_T1  affected         t1
A3_T3  affected         t3
A4_T1  affected         t1
A4_T2  affected         t2
A5_T1  affected         t1
A5_T3  affected         t3
A6_T1  affected         t1
A6_T2  affected         t2
W1_T1   control         t1
W1_T2   control         t2
W1_T3   control         t3
W2_T1   control         t1
W2_T2   control         t2
W2_T3   control         t3
W3_T1   control         t1
W3_T2   control         t2
W3_T3   control         t3
W4_T1   control         t1
W4_T2   control         t2
W4_T3   control         t3

I used to separate out t1, t2, and t3 samples in the DESeqDataSetFromMatrix step to make three separate. Recently, I read the explanation for "If I have multiple groups, should I run all together or split into pairs of groups? " So, I would like to run samples from all groups together, and then use the contrast argument of the results function to extract comparisons of interest after fitting the model using DESeq.

If I simply used 

deseq_out <- DESeq(deseq)
deseq_de_out<-results(deseq_out, contrast = c("condition","affected", "control"), alpha = 0.05)

Then the result would be comparing all affected samples (t1, t2, and t3) all together with all control samples. This is not what I want. 

Is it possible to compare just affected at t1 with control at t1 with the contrast function? Thanks!

Can you take a look at our time course example in the workflow, and see what question you have remaining?

http://www.bioconductor.org/help/workflows/rnaseqGene/#time