Closed royfrancis closed 2 years ago
For whatever analysis you're going to do after SoupX, you'll need to know which droplets contain cells. Once you have decided which droplets contain cells, then you should use this to construct the table of counts toc
object and pass that (along with the raw counts as tod
) to SoupChannel
to construct the object needed by SoupX.
You are free to identify which droplets contain cells however you like. One option is to use cellranger, but you could use something like the DropletUtils R package or just use a cut-off in number of UMIs.
I only have raw_feature_bc_matrix. Is it possible to run this tool? Or should I rerun CellRanger and generate the filtered counts?