Hi,
I am following a DESeq2 tutorial (see link below) and I have a question about setting up the experiment table and design formula.
The data that I have is made up of 3 groups (control, knockout_a, knockout_ab), and each group has 4 replicates.
Here is the experiment design table that I created:
| run | knockout | |
| control_replicate_1 | control_replicate1 | control |
| control_replicate_2 | control_replicate2 | control |
| control_replicate_3 | control_replicate3 | control |
| control_replicate_4 | control_replicate4 | control |
| a_replicate1 | a_replicate1 | a |
| a_replicate2 | a_replicate2 | a |
| a_replicate3 | a_replicate3 | a |
| a_replicate4 | a_replicate4 | a |
| ab_replicate1 | ab_replicate1 | ab |
| ab_replicate2 | ab_replicate2 | ab |
| ab_replicate3 | ab_replicate3 | ab |
| ab_replicate4 | ab_replicate4 | ab |
I am trying to find find the differential expression of:
control vs knockout_a
control vs knockout_ab
knockout_a vs knockout_ab
I would also like retain the ability do PCA for all 3 groups (1 plot) and for each comparison group (3 plots).
When I run ( dds <- DESeqDataSet(se, design = ~ knockout) ) I believe it just compares "control vs ab".
Is there a way to modify the design table and/or design formula to express what I am trying to do? Or will I have to create 3 separate design tables (for each of the 3 comparisons) to achieve my goal? Along with that, any clear explanation for how the design table and design formula are connected would be much appreciated.
Thanks for your help!
