Wrapper scripts for components of the DropletUtils package

DropletUtils is a Bioconductor package written by Aaron Lun. This package provides parameterised scripts providing direct command line access to the functions of that package, to facilitate inclusion in workflows.

Install

The recommended method for script installation is via a Bioconda recipe called dropletutils-scripts.

With the Bioconda channels configured the latest release version of the package can be installed via the regular conda install command:

conda install dropletutils-scripts

Test installation

There is a test script included:

dropletutils-scripts-post-install-tests.sh

This downloads test data and executes all of the scripts described below.

Commands

Currently available scripts are detailed below, each of which has usage instructions available via --help.

dropletutils-read-10x-counts.R: call Read10xCounts()

To read 10x data and create a SingleCellExperiment object comprising sparse matrices:

dropletutils-read-10x-counts.R -s <comma-separated list of 10x directories> -c <logical, should columns be named by barcode?> -o <file to store serialized SingleCellExperiment object>

dropletutils-downsample-matrix.R: call downsampleMatrix()

dropletutils-downsample-matrix.R -i <input SingleCellExperiment in .rds format> -p <single proportion or file with one value per line for each matrix column> -c <logical, should downsampling be done by column?> -o <file to store serialized SingleCellExperiment object>

dropletutils-barcoderanks.R: call barcodeRanks()

dropletutils-barcoderanks.R -i <input SingleCellExperiment in .rds format> -l <maximum total count below which all barcodes are assumed to be empty droplets> -f <fit bounds> -o <file to store serialized SingleCellExperiment object> -p <file to store barcode plot in>

dropletutils-empty-drops.R: call emptyDrops() and produce filtered object if desired

dropletutils-empty-drops.R -i <input SingleCellExperiment in .rds format> -l <maximum total count below which all barcodes are assumed to be empty droplets> \
    -n <number of iterations> -m <logical, return barcodes with totals less than -l value?> -g <minimum total count required for a barcode to be considered as a potential cell> \
    -r <retain> -f <logical, filter the output object?> -d <FDR value below which a droplet is said to contain a cell> -o <file to store serialized SingleCellExperiment object> -t <output text file>