Correct for different number of cells and individuals in single cell data analysis

Question

[Sara]

I have scRNAseq from different donors for 4 different conditions. for different condition I have different number of donors and for different condition we have different number of cells. to make the conditions comparable, is there any way to correct for these differences?

[ATpoint]

What analysis do you have in mind? Is pseudobulking an option, like, summing counts of cells per gene and cluster/donor/celltype, whatever the setup is?

[Sara]

yes. I am planning to do DGE analysis.

[ATpoint]

...and does the design allow pseudobulks? That would be preferred, both in terms of statistics, as you do not have the pseudoreplication and inflated power issue of using single-cells, and not the issue of different power due to different cell numbers.