run_cellranger.pl

Create and run a snakemake pipeline that launches cellranger commands to biowulf.

View the entire readme with run_cellranger.pl -h

Parts of the cellranger workflow

This script produces a Snakemake file that will invoke a series of Cellranger commands to analyze single-cell RNA-seq data. The Cellranger steps invoked are:

1. mkfastq
2. count
3. mat2csv

Additionally, Seurat objects will be made for each sample.

Preparing for the run

Input consists of a tab-delimited file with the following column order, with each row representing a sample:

1. Library ID
2. Sample ID [must be unique]
3. Raw Data Directory
4. Lane
5. Sample Barcode Index
6. Group

'Sample ID' must be unique. 'Group' determines what samples will be merged together during the Seurat merge step. All samples with a given Group ID will be merged for that group. Samples can be in more than one group (separate Group ID values with commas), or can be ignored during merging by setting this field to 'NA'.

TODO:

1. [ ]Cleanup intermediate files (could be done using native snakemake nomenclature)

Author

Jason Inman inmanjm@nih.gov