immunopipe 0.5.0

A pipeline for integrative analysis for scTCR- and scRNA-seq data

immunopipe

Integrative analysis for scTCR- and scRNA-seq data

Requirements & Installation

• python: 3.7+

  • Other python depedencies should be installed via pip install -U immunopipe

• R

  • A bunch of R packages

• Other

  • VDJtools: https://vdjtools-doc.readthedocs.io/en/master/install.html

• Checking requirements

  pip install -U pipen-cli-require
  pipen require immunopipe.pipeline:pipeline <pipeline arguments>
  

• Quick way to install the dependencies using conda

  conda env install --name <env_name> --file docker/environment.yml
  # then
  conda activate <env_name>
  

Running as a container

Using docker:

docker run \
    -w /immunopipe/workdir \
    -v $(pwd)/:/immunopipe/workdir \
    -v /tmp \
    -v $(pwd)/prepared-data:/mnt \
    justold/immunopipe:<tag>  # or :dev to use the development version

Using singularity:

singularity run -w \  # need it to be writable
  --pwd /immunopipe/workdir -B .:/immunopipe/workdir \  # Could use other directory instead of "."
  # --contain: don't map host filesystem
  # --cleanenv: recommended, to avoid other host's environment variables to be used
  #   For example, $CONDA_PREFIX to affect host's conda environment
  --contain --cleanenv \
  docker://justold/immunopipe:<tag>  # or :dev to use the development version

# The mount your data directory to /mnt, which will make startup faster
# For example
#   -B .:/immunopipe/workdir,/path/to/data:/mnt
# Where /path/to/data is the data directory containing the data files
# You may also want to bind other directories (i.e. /tmp)
#   -B <other bindings>,/tmp

# Or you can pull the image first by:
singularity pull --force --dir images/ docker://justold/immunopipe:<tag>
# Then you can replace "docker://justold/immunopipe:<tag>" with "images/immunopipe.sif"

Modules

• Basic TCR data analysis using immunarch
• Clone Residency analysis if you have paired samples (i.e. Tumor vs Normal)
• V-J usage, the frequency of various V-J junctions in circos-style plots
• Clustering cells and configurale arguments to separate T and non-T cells
• Clustering T cell, markers for each cluster and enrichment analysis for the markers
• Radar plots to show the composition of cells for clusters
• (Meta-)Markers finder for selected groups/clones of cells
• Psedo-bulk GSEA analysis of two groups of cells
• Seurat cluster statistics, including:
  • Basic statistics of the clusters (e.g. number of cells in each cluster)
  • Gene expressions (e.g. ridge, violin, feature, dot and heatmap plots)
  • Dimensional reduction plots
• TCR clustering using CDR3 sequences and the statistics of the clusters
• Cell group distribution (TCR clones/clusters) in Seurat clusters
• Clone heterogeneity (TCR clone distribution) in Seurat clusters
• Metabolic landscape analysis (Ref: Xiao, Zhengtao, Ziwei Dai, and Jason W. Locasale. "Metabolic landscape of the tumor microenvironment at single cell resolution." Nature communications 10.1 (2019): 1-12.)

Documentaion

https://pwwang.github.io/immunopipe

Example

https://github.com/pwwang/immunopipe-example