- System requirements
- All software dependencies and operating systems (including version numbers) are available in the methods section
- Archived version of this repo has been tested on x86_64-pc-linux-gnu (64-bit) platform running under Ubuntu 22.04.1 LTS with R version 4.2.2 and Python version 3.8.8
- GPU is required to run ambient RNA removal with CellBender
- Installation guide
- Installation is available via Snakemake framework infrastructure. The Snakefile pipeline is in the project root. Before run: 1) You need to edit path to the cellranger environment file on your computer ($CELLRANGER). 2) Be sure that Apptainer (former Singularity is installed and available in $PATH).
- Typical install time on a "normal" desktop computer should be within a hour that you need to install cellranger, Anaconda, Apptainer/Singularity, to create snakemake conda environment and to download all configured container-images.
- Demo
- We include preprocessed data that one can use to reproduce panels of the figure 2 related to snRNA-seq data from the manuscript.
- It is expected that output images can slightly differ due to random factor native to dimension reduction technics.
- Expected run time for demo on a "normal" desktop computer is within a hour.
- Instructions for use
- To run the demo on preprocessed data you need to source code/render.R file, which will rerun rendering of Rmarkdown analysis notebooks for figure 2 via workflowr package.
- To reproduce whole pipeline you would need to modify and run command
snakemake -f --use-singularity -prk --cores <FILL-IN-N-CORES> --resources nvidia_gpu=1 --singularity-args "--nv --bind <FILL-IN-PATH-TO-CLONED-REPO>" --rerun-incomplete --retries 1 --resources mem_mb=<FILL-IN-MEM>