I have 16 samples and 2-factor conditions (treatment and sex):

sample treatment sex group

1 untreated male untreated_male

2 untreated male untreated_male

3 untreated male untreated_male

4 untreated male untreated_male

5 untreated female untreated_female

6 untreated female untreated_female

7 untreated female untreated_female

8 untreated female untreated_female

9 treated male treated_male

10 treated male treated_male

11 treated male treated_male

12 treated male treated_male

13 treated female treated_female

14 treated female treated_female

15 treated female treated_female

16 treated female treated_female

I want to do 3 types of DEG analysis using DESeq2.

1. Treated_female VS Untreated Female
2. Treated_male VS Untreated male
3. Treated VS Untreated controlled by sex

I read in a post that was similar to my design that for the #1 and #2 analysis, it is better to combine all the 16 samples in dds and then set a design just specifically for say treated_female VS untreated Female.

I have created one other group so in total I have 3 cols: treatment, sex, group (condition_group).

Then I have following design:

dds <- DESeqDataSetFromHTSeqCount( sampleTable = sampleTable, directory = directory, design= ~group)

dds <- DESeq(dds)

contrastmale <- c("group", "treatedmale", "untreated_male")

resmale <- results(dds, contrast = contrastmale, alpha = 0.05)

contrastfemale <- c("group", "treatedfemale", "untreated_female")

resfemale <- results(dds, contrast = contrastfemale, alpha = 0.05)

class(res_male)

resultsNames(dds)

[1] "Intercept"

[2] "grouptreatedmalevsuntreated_female"

[3] "groupuntreatedfemalevstreated_female"

[4] "groupuntreatedmalevstreated_female"

res_male

log2 fold change (MAP): group treatedmale vs untreatedmale

Wald test p-value: group treatedmale vs untreatedmale

DataFrame with 21971 rows and 6 columns

My question is, why I do not have the right comparisons in the resultsName ( it does not even have treatedmale vs untreatedmale)? even though the resmale shows that the comparison is treatedmale vs untreated_male

Don’t worry about the coefficient names printed by resultsNames(). When you use the contrast argument the software composed the correct comparison for you. You can look at plotCounts of the top genes by pvalue to convince yourself.