CLI utilities¶

Besides running the pipeline, immunopipe ships a set of utility commands for inspecting the pipeline outputs after a run:

$ immunopipe utils --help

check-genes¶

Verify that gene symbols exist in the data before visualizing their expressions in SeuratClusterStats. The Seurat object is read from the output of the SeuratPreparing process.

$ immunopipe utils check-genes -w ./pipen/<pipeline name> -g CD3D,CD4,CD8A

Options¶

• -w, --workdir (required): Working directory of the pipeline, typically ./pipen/<pipeline name>
• -g, --genes (required): Comma-separated gene symbols to check, or a file path prefixed with file:// containing gene symbols (one per line). Tab-separated files are also accepted, in which case only the first column is used.
• --assay: The assay to check the genes against (e.g. RNA, ADT). The genes are checked against the default assay if not specified.
• --rscript: Path to the Rscript executable (default: Rscript)

Example¶

# Check comma-separated genes
$ immunopipe utils check-genes -w ./pipen/pipeline1 -g CD3D,CD4,CD8A

# Check genes from a file
$ immunopipe utils check-genes -w ./pipen/pipeline1 -g file://genes.txt

# Check genes in a specific assay
$ immunopipe utils check-genes -w ./pipen/pipeline1 -g CD3D,CD4 --assay ADT

check-dim¶

Print the cell and gene counts before and after QC to inspect the effect of the QC filtering in SeuratPreparing. The output tables are the full contents of qc/cell_qc.txt and qc/gene_qc.txt in the SeuratPreparing output directory.

$ immunopipe utils check-dim -w ./pipen/<pipeline name>

Options¶

• -w, --workdir (required): Working directory of the pipeline, typically ./pipen/<pipeline name>
• --rscript: Path to the Rscript executable (default: Rscript)

select-markers¶

Select the top marker genes for each cluster from the ClusterMarkers process, for downstream analysis and visualization. The selected markers are written to stdout as a tab-separated table, so you can redirect it to a file:

$ immunopipe utils select-markers -o ./pipen/<pipeline name> > selected_markers.tsv

Options¶

• -o, --outdir (required): Output directory of the pipeline (the directory containing the ClusterMarkers output)
• -t, --top-n: Number of top markers to keep per cluster (default: 10)
• --order-by: An R expression to order the markers by before selecting the top ones. Available variables: avg_log2FC, pct.1, pct.2, p_val, p_val_adj (default: desc(avg_log2FC))
• -f, --filter: An R expression to filter the markers before selection. Available variables: avg_log2FC, pct.1, pct.2, p_val, p_val_adj (default: p_val_adj < 0.05)
• --rscript: Path to the Rscript executable (default: Rscript)