Hi,

Apologies if this question doesn't fit the scope of the forum.

How many replicates (biological/technical) do you need when doing single cell RNA-seq using cell lines?

We are planning to do a pilot study using scRNA-seq on cancer cell lines. We will use 4 conditions (3 treatments + control) and at two timepoints. That is 8 samples "per replicate". We have been advised to use the 10x Flex assay multiplexing 8 samples in a single run.

We have done a similar experiments before, using bulk RNA-seq, and there we used 3-4 replicates per condition. However, in scRNA-seq going from 2 to 3 replicates would increase the budget required by 50%, and that would go above our current budget.

Another option would be to remove 1 of the treatments of the analysis, and then having 3 replicates of the 2 treatments x 2 timepoints (3x2x2 = 12 samples), but only 2 replicates of the controls (2x2timepoints = 4 samples). With 12+4 = 16 fitting in 2 sequencing runs multiplexing 8 samples on each.

I'm asking because we expect to obtain ~10k cells per sample. I'm not sure if adding a 3rd replicate of the controls with 10k more cells with no response to any treatment will actually improve the power of the analyses, or it would be a mere formality and a waste of money.

Do you have any insights on this?

Thanks