Hi I'm trying to run differential expression using DESeq2 and I'm a little confused about how to interpret results based on contrast. In my dataset I'm looking at 2 levels within my variable "stage" (B and R) and I want to visualize the results on a volcano plot.

I have created my DESeq object using the following code:

    ddsWald_stage <- DESeqDataFromMatrix(countData = countData, colData = colData, design ~= stage)

I then created my results object using:

res_stage <- results(ddsWald_stage, contrast = c("stage", "B", "R"), cooksCutoff=TRUE)

When I plot this on a volcano plot, will genes that have significantly higher fold changes in the B group be on the right or left side of the plot? I'm a little confused by the documentation for contrast. Any help would be appreciated!

In this case, the comparison will be conducted with B as the numerator, and R as the denominator, i.e., B / R (B divided by R). It therefore follows that genes having positive fold changes will be higher in group B when compared to group R, and these will indeed appear on the right hand side of the volcano plot.

If you don't specify the order for the comparison, then the order will be determined based on the ordering of the levels in the stage variable. In R, one can re-order factor levels like this:

colData$stage <- factor(colData$stage, levels = c('R', 'B')

...or:

colData$stage <- relevel(colData$stage,  ref = 'R')

Both of these will set R as the reference level.

The Kevin