Open showteeth opened 3 years ago
Just a thought: You could use a package like Seurat in R to integrate the datasets, cluster all the cells together, and label cell types. Not sure if the statistical assumptions would be broken in CellPhoneDB, but you could then concatenate the raw count matrices and use that (plus the metadata) as the inputs to CellPhoneDB. You want to use the raw counts and not the integrated features, as this would otherwise violate statistical assumptions in the statistical _analysis function.
Hi,
I have data obtained from two different single-cell RNA-seq platforms:
Now, I wonder if I can apply cellphonedb to above datasets to get cell-cell communications.
Thanks!