DESEQ2 control vs treat based on alphabetic order?

[FelipeAd]

Hello,

I am using DESeq2 for differential expression analysis of RNA Seq data.

My question is: which is considered "control" or "treated" in the "countData" and "condition" table.

Let me explain.

When having "countData" table in the format of "control" "control" "control" "control" "treated" "treated" "treated" "treated" table, a also assign the "condition" as follows:

condition <- factor(c(rep("control", 4), rep("treat", 4)))

Following analysis, I reach to the

res <- results(dds)

and
head(res) gives

log2 fold change (MAP): condition treat vs ctrl
Wald test p-value: condition treat vs ctrl

The exact same result I get when I reverse the "condition":

condition <- factor(c(rep("treat", 4), rep("control", 4)))

or when I reverse the countData.

But when I set

condition <- factor(c(rep("atreat", 4), rep("control", 4)))

my head(res) gives

log2 fold change (MAP): condition ctrl vs atreat
Wald test p-value: condition ctrl vs atreat

So it seems that the vs comment is based on alphabetic order?

[dpryan]

Factors are ordered alphabetically in R unless you specify the order yourself:

condition <- factor(c(rep("WT",4),rep("Mut",4)), levels=c("WT","Mut"))

Have a look at:

levels(condition)

[Michael Love]

It's not DESeq2, but R's base function 'factor' which determines the order of levels in a factor.

You can control the order of the levels either using the 'contrast' argument, or while you are setting up the factors.

If you read over the DESeq2 vignette:

vignette("DESeq2")

We have in the beginning:

"1.2.5 Note on factor levels

In the three examples above, we applied the function factor to the column of interest in colData, supplying a character vector of levels. It is important to supply levels (otherwise the levels are chosen in alphabetical order) and to put the “control” or “untreated” level as the first element (”reference level”), so that the log2 fold changes produced by default will be the expected comparison against the reference level, that is log2 (treated/untreated). An R function for easily changing the reference level is relevel. An example of setting the reference level of a factor with relevel is..."

[FelipeAd]

Thank you both.

The:

condition <- factor(c(rep("WT",4),rep("Mut",4)), levels=c("WT","Mut"))

solved the problem.