Hello, I am doing a scRNA-Seq analysis and I want to use doubletCells() function to identify possible doublets. My data comes from 4 different batches and I use fastMNN for batch correction. Which way would be better in this situation? Using doubletCells() for each data before batch correction and remove cells with high scores as doublets and doing the batch correction or after fastMNN(), using the doubletCells() function on counts of all data sets (I think I have to use computeSumFactors() function with counts of all data sets).
Thank you in advance.