DESeq2 design formula with paired samples

Question

[helen]

I have 6 subjects 1-6, and would like to identify the differentially expressed genes in condition A vs.B and considering the gender effect.

Here is my coldata:

subject gender  condition
      1      m          A
      1      m          B   
      2      m          A
      2      m          B
      3      f          A
      3      f          B
      4      f          A
      4      f          B
      5      m          A
      5      m          B
      6      f          A
      6      f          B

I think the condition is nested in the subject, right? But when I tried the following design formula, DESeq2 returned Error in checkFullRank(modelMatrix)

dds <- DESeqDataSetFromMatrix(countData = acts,
                              colData = coldata,
                              design = ~ subject + gender + condition + condition:subject)

Although the following combinations can be processed:

design = ~ subject + condition or

design = ~ condition + gender

But how do I modify my design formula to consider paired samples, gender effects, and conditions (or any interaction) at the same time? Thanks!

Answer 1

[helen]

My updated post is here: How to interpret DESeq2 result?