search

sqjin / CellChat

R toolkit for inference, visualization and analysis of cell-cell communication from single-cell data
GNU General Public License v3.0
623 stars 142 forks source link

Subsets of Seurat Object and permutation testing in CellChat #425

Open HaTonemann opened 2 years ago

HaTonemann commented 2 years ago

Hi CellChat Team. First of all, thank you for the amazing package and the detailed tutorial, which makes it fun to work with CellChat!

Since many CellChat users seem to use subsetted Seurat objects for CellChat analysis - especially for computeCommunProb() which uses permutation testing in behind - I wonder whether it is cleaner to use a subset of clusters of a Seurat object in advance (only the clusters of interest), or to infer the cell-cell interactions on all clusters of the Seurat object and subset afterwards?

I am not sure whether there is a right or wrong, I am only interested in your opinion on the two workflows, since the amount of clusters probably affects the results of the permutation testing.

Thanks in advance!

sqjin commented 2 years ago

@HaTonemann It is true that the amount of clusters may affect the results of the permutation testing. I think you can subset the cell clusters based on the biological questions you want to answer.