Minimum Number of Cells for Downstream Analysis after subsetting a particular cell type (scRNAseq)?

Question

[Tundup]

Suppose if I have two groups and i want to look into particular subtypes say Alveolar Macrophages.

What should be the minimum number of cells for the downstream analysis? I can see many reasons and numbers in the literature.

Please give reasons for your answer.

Answer 1

[ATpoint]

There is no answer to this. Some people will say 10, other 3, others 500. Generally, since single-cell is noisy and has dropouts, even for canonical marker genes, you want some cells to be sure that you can redundantly find expression in not just a handful of cells. Marker detection or differential expression is generally poor on single-cell level, and even more so if you just have a couple of cells. I don't think there is this one single answer to satisfy you. You also don't give any details what "look into" means.