Description of feature
Currently, cellbender is used to remove empty droplets from the input anndata. This run of cellbender also produces a filtered anndata file with ambient rna removed.
However, if you set cellbender as the method to remove ambient rna, it gets run again from scratch. The two runs are not identical, as the second run bases the priors on the first run, but I feel that this is quite redundant and perhaps the result from the first run is sufficient for empty droplet removal, as well as ambient rna filtering.
I might be missing something that happens in the background, but I think that this might just be wasting resources.
Description of feature
Currently, cellbender is used to remove empty droplets from the input anndata. This run of cellbender also produces a filtered anndata file with ambient rna removed.
However, if you set cellbender as the method to remove ambient rna, it gets run again from scratch. The two runs are not identical, as the second run bases the priors on the first run, but I feel that this is quite redundant and perhaps the result from the first run is sufficient for empty droplet removal, as well as ambient rna filtering.
I might be missing something that happens in the background, but I think that this might just be wasting resources.