Hello,

I am using deseq2 to understand differentially expressed genes in a asthma study. I wanted to know the interaction of treatment with Sex.

My sample table looks like (condition relates to treatment):

sampleName         condition       Sex    

S-001240813         Sham    M       

S-001240814        Dex     M      

S-001240815       Sham    F      

S-001240816       Dex     F       

S-001240817       Sham    M      

S-001240818       Dex     M       

S-001240819       Sham    F       

S-001240820      Dex     F       

S-001240821      Sham    F      

S-001240822       Dex     F       

S-001240823      Sham    F      

S-001240824       Dex     F       

S-001240825     Sham    M       

S-001240826     Dex     M      

S-001240827     Sham    M       

S-001240828     Dex     M       

S-001240829     Sham    F       

S-001240830     Dex     F      

S-001240831     Sham    F       

S-001240832     Dex     F       

S-001240833     Sham    M       

S-001240834     Dex     M       

S-001240835      Sham    F       

S-001240836      Dex     F     

S-001240837      Sham    M      

S-001240838      Dex     M       

I used design string ~Sex + condition + Sex:condition

But when I checked my object after running deseq2 it only gave me one interaction out of four

> resultsNames(dds)
[1] "Intercept"             "Sex_M_vs_F"            "condition_Sham_vs_Dex"
[4] "SexM.conditionSham"   

Can you please help figure out my mistake

This is correct. There is only a single interaction term in this 2x2 model. A result table for this term tests whether the condition effect is equal across sex (small p-value suggests rejection of the null hypothesis that the effect is equal). See the details in the vignette section on Interaction terms.

If you have further questions beyond the explanation in the vignette, I'd recommend consulting either a statistics textbook or a local statistician who can help explain interaction terms in linear models.