Have you used ComBat for adjusting batch effects using an empirical Bayes framework ? If yes, I have few doubts:

1) Why we should use this ?

2) If we want to process data for DEG analysis using DESeq2, EdgeR, etc. should we go for ComBat ?