Hi, This is my first time using the DESeq2 package for RNA-seq analysis. I am analyzing a study of three different time points in mice (5,10 and 15 days old), all mice with the same genotype, and 3 biological replicates per age. I want to evaluate differential expression as an effect of the different ages. I am confused on how I should correct for the variability associated to the biological replicates when checking for the effect of age.

I have tried both these options for the code but I always get very small p-values (mostly 0) and very high number of genes differentially expressed (half of the detected genes), which I don't think makes any biological sense. ``` Option 1:

sampleTable <- data.frame(condition = relevel(factor(rep(c("P10", "P15", "P5"), each = 3)), ref='P5'), rep1= factor(rep(c("1","2","3"), times=3))

dds <- DESeqDataSetFromMatrix(countData = countdata,

colData=sampleTable,

design= ~condition + rep 1)

Option 2:

sampleTable <- data.frame(condition = relevel(factor(rep(c("P10", "P15", "P5"), each = 3)), ref='P5'), rep2= factor(rep(c("1","2","3","4","5","6","7","8","9"), times=1)))

dds <- DESeqDataSetFromMatrix(countData = countdata,

colData=sampleTable,

design= ~condition + rep2),

Any help with how to properly use "design" would be greatly appreciated!

For you it would simply be:

DESeqDataSetFromMatrix(countData=countdata,
                       colData=data.frame(condition = factor(rep(c("P10", "P15", "P5"), each=3))),
                       design="~condition")

Each element of colData with the same factor is considered a replicate, there is no need for (it would even be wrong) something like ~rep.