A Nextflow workflow for single-cell RNA-seq cell type annotation using Azimuth reference datasets.
This pipeline performs automated cell type annotation of single-cell RNA-seq data using the Azimuth algorithm. It can process data against reference datasets and optionally generate interactive visualizations using Vitessce.
- Nextflow (≥21.04.0)
- Docker or Singularity
- Reference Files (use Zenodo link to download)
nextflow run main.nf \
--input /path/to/input.h5 \
--reference /path/to/reference_folder \
--output /path/to/output/
nextflow run main.nf \
--input /path/to/input.h5 \
--reference /path/to/reference_folder \
--output /path/to/output/ \
--web_output /path/to/web/output/
Parameter | Description | Example |
---|---|---|
--input |
Input single-cell data file (H5 Matrix or H5Seurat object) | /data/sample.h5seurat |
--reference |
Azimuth reference dataset file path | /data/human_pbmc |
--output |
Output directory for results | /results/ |
Parameter | Description | Default | Example |
---|---|---|---|
--web_output |
Directory for web visualization files | false |
/web/output/ |
--name |
Analysis name | "Azimuth Analysis" |
"My Analysis" |
--description |
Analysis description | "Mapping of gene expression data onto Azimuth reference" |
"Custom description" |
*.h5seurat
- Annotated single-cell data in H5Seurat format*.h5ad
- Annotated single-cell data in H5AD format (if conversion enabled)
vitessce.config.*.json
- Vitessce configuration filechart.manifest.json
- Chart manifest for Cirro visualizationazimuth.zarr/
- Zarr-formatted data for web viewing