Hello all,

I am analyzing an RNAseq with 6 conditions using deseq2. I have 6 biological replicates for control, however for the other 5 conditions (say C1,C2,C3,C4,C5) I have no replicates. Ideally, I would like to do a variance estimation using the 6 replicates for the control, and using it compare any two condition C1 vs C2. Does the standard workflow takes care of this?

thanks in advances

The standard workflow will work (the software will run and it will give you results for C1 vs C2 comparisons, etc. This model will assume that the within-condition variation in the control samples is similar to the conditions. This may not be the case, which is why biological replicates for all conditions is preferred, to get a better estimate for the variation across all groups. Additionally biological replicates for the condition samples would help in reducing the noise in the estimate for the condition effect size and direction. With only one sample each, noise might result in the wrong size of direction of fold change.

duplicated text.....