To run this workflow, you must supply the following parameters:
- Sample: The name of your sample. Output files will be have this sample as a prefix.
- Input: An .h5 file containing the raw (not filtered) cell-by-gene matrix. For experiments using the 10x Genomics platform, this file can be generated using the Cell Ranger software package.
- Expected cells: The number of cells expected in this experiment. An estimate within a factor of 2 is sufficient for this value. If left at 0 (the default), CellBender will estimate this value.
- Total droplets included: The total number of droplets for CellBender to infer cell probabilities. If left at 0 (the default), CellBender will estimate this value.
- Sample sheet: A CSV file containing a header with the following fields:
- Sample: A unique name of the sample.
- Location: The name of the .h5 file containing a **raw** (not filtered)
count matrix.
- Expected_cells: The number of cells expected in this sample. This
value need only be accurate to within about 2x. Set this to 0 for
autodetection (recommended).
- Total_droplets_included: The number of droplets from the barcode-rank
plot that could contain true cells. Any droplets beyond this cutoff
will be considered empty droplets. Set this to 0 for autodetection
(recommended).
- Input: Select the .h5 files which are outlined in the Location field in the
sample sheet.
See the next tab for information on outputs and how to use them.