ComBat-seq adjustment of batch effects

Question

[chatterjee.arumoy]

Hi,

If I have done my RNA seq with 4 samples X 3 replicates each in two batches (see below for the design matrix):

1) Is it correct to adjust batch effects on the raw count file first by ComBat-seq (Zhang et.al., 2020) and then process the output by edgeR?

CONDITIONS  SAMPLES BATCH
T1            R1      1
T1            R2      1
T1            R3      1
T2            R1      1
T2            R2      2
T2            R3      2
T3            R1      1
T3            R2      1
T3            R3      1
T4            R1      1
T4            R2      2
T4            R3      2

I am asking this question since the only R2 and R3 for conditions T2 and T4 were processed in batch 2 and I could not include such a batch effect in the edgeR model matrix (see below) which needs batch 2 to be present in each of T1, T2, T3 and T4.

design = model.matrix(~ 0 + conditions + batch, data=y$samples)

Ref: Zhang et.al., 2020: https://academic.oup.com/nargab/article/2/3/lqaa078/5909519

Kindly help.

Thanks

Arumoy.