Hello, I have a question about how to design a formula.

I have the following samples with three factors, Names, sample and condition. A total of 18 samples, divided into 9 groups(A-I), I want to use DESeq2 to analysis paired samples, like difference of sample A before and after treatment(trt vs untrt).

 Names sample condition
X1  1   A      untrt
X2  2   A       trt
X3  3   B      untrt
X4  4   B       trt
X5  5   C     untrt
X6  6   C       trt
X7  7   D     untrt
X8  8   D       trt
X9  9   E     untrt
X10 10  E       trt
X11 11  F     untrt
X12 12  F       trt
X13 13  G     untrt
X14 14  G       trt
X15 15  H     untrt
X16 16  H       trt
X17 17  I     untrt
X18 18  I       trt

I am not sure how to design the formula. Is that right?

dds_1 <- DESeqDataSetFromTximport(txi_import, colData = sampleTable, design = ~ sample + condition)

Hi xiaqiu0809,

Yes, this represents a paired design. So, once you derive your test statistics for condition, these will be controlled for the sample pairing. Please ensure that sample is encoded categorically (as a factor)..

There is a previous answer here, to which I will link you: https://support.bioconductor.org/p/84241/#84256

Kevin