tiger-snake

Overview

Pipeline for analysis of single cell sequencing data of RNA and T-cell receptors, examining TCR-seq (t), immune cell annotation (i), gene expression (g), epitope annotation (e), and repertoires of clones (r).

---

Modules

The Snakefile reads from the input configuration file (config template: config/config.yaml).

processing

The processing module prepares single cell RNA-seq data for analysis in the following steps:

1. make individual Seurat objects of samples
2. filter and process individual sample Seurat objects

integration

The integration module prepares single cell RNA-seq data for analysis in the following steps:

1. merge together individual Seurat objects
2. run integration (using Harmony) on merged Seurat object

annotation

The annotation module performs the following annotations of the Seurat object:

1. run SingleR to annotate the Seurat object
2. add new metadata from input sample sheet
3. relevel cell types (so they will plot in a certain order)
4. add SingleR, metadata, and cell type releveling info to Seurat object
5. add TCR data to Seurat object

analysis

The analysis module performs the following analyses of the Seurat object:

1. differential expression
2. Chi-square tests
3. TCR epitope annotation
4. TCR clontype overlap analysis

---

Environment set up

The environment set up for tiger-snake is a conda environment with Snakemake (v9.17.2) installed.

The tiger-snake pipeline is configured to set up the conda environment in snakemake/envs/conda_setup.sh (which is souced when running the main controller script, tiger-snake.sh).

In the conda set up file ([snakemake/envs/conda_setup.sh]), the versions of Snakemake and Nextflow can be changed if different versions of the software are desired.

---

Pipeline execution

Example command

The pipeline can be run by invoking the main bash script, tiger-snake.sh.

bash tiger-snake.sh \
  --config config/config.yaml \
  --runtype dry-run \
  --executor slurm

Arguments

The main controller script is tiger-snake.sh. The following argument options are:

Flag	Required?	Description	Options
--config, -c	✓	Path to config file	—
--runtype, -r	✓	How Snakemake is invoked	run dry-run
--executor, -e	✓	Executor/profile mode	slurm local

Detailed explanation of flags:

• --config, -c (required): Path to config file (see config/config.yaml for the required structure)

• --runtype, -r (required): How Snakemake is invoked

  • runtype — executes normally
  • dry-run — passes --dry-run flag to Snakemake

• --executor, -e (required): Executor mode / profile selection

  • slurm — enables slurm executor with slurm directives
  • local — runs pipeline on the local machine

---

Development

The tiger-snake pipeline is still being developed.

Software versions used for development and testing

Base software

• Snakemake v9.17.2
• R v4.5.3
  • Bioconductor v3.23
  • see renv.lock for all versions of R packages