I was wondering what the best way to compare data from separate single cell experiments was?
I am hoping to be able to compare a dataset containing ~30 cells of the same type, to another (heterogeneous) single cell dataset, where each cell has a much lower sequencing depth. One or two of the cell types from the second dataset are of particular interest.
I was wondering what packages would be best to perform this? I'm assuming if I had access to the full dataset I could potentially just subset the matrix to include only cells from the populations I wanted to compare to (assuming I knew which cells were which).
I had a read through of this workflow, and am wondering if either mnnCorrect from the scMap package, or CCA from Seurat will help me achieve this. If so, would I be able to use either of these to compare the original ~30 cells to specific clusters in the second dataset, i.e. via differential expression analysis? I see that in the case of mnnCorrect, the data will no longer be stored as counts and may contain negative values which may prevent certain analysis tools being used.
Any advice would be greatly appreciated.