Hi there,

I am performing RNA-seq analysis on some CD4 T cells from 3 different patients, each treated or untreated with a drug. For using the DESeqDataSetFromTximport function, I am wondering the best way to do this.

My samples are as follows:

> sampleID2
     sampleID         conditions patientID
[1,] "1"                   "untreated"     "A"  
[2,] "2"                   "treated"      "A"  
[3,] "3"                   "untreated"     "B"  
[4,] "4"                   "treated"      "B"  
[5,] "5"                   "untreated"     "C"  
[6,] "6"                   "treated"      "C"  

> as.data.frame(sampleID2) -> sampleID2
> as.factor(sampleID2$conditions) ->sampleID2$conditions
> as.factor(sampleID2$patientID) ->sampleID2$patientID
> ddsTxi2 <- DESeqDataSetFromTximport(txi,
+                                    colData = sampleID2,
+                                    design = ~ patientID + conditions)

So my question is, should I be doing "design = ~ patientID + conditions" or just ignoring the patient information and treat the experiment like 3 replicates with two conditions, "treated" and "untreated" and do "design = ~ conditions"?

I would just like to know what genes are differentially expressed between treated and untreated cells, but want to make sure I am doing this the best way.

Thanks so much for your help!

Luisa