This repository holds a Nextflow workflow (scpca-nf) that is used to process 10X single-cell data as part of the Single-cell Pediatric Cancer Atlas (ScPCA) project.
All dependencies for the workflow outside of the Nextflow workflow engine itself are handled automatically; setup generally requires only organizing the input files and configuring Nextflow for your computing environment.
Nextflow will also handle parallelizing sample processing as allowed by your environment, minimizing total run time.
The workflow processes fastq files from single-cell and single-nuclei RNA-seq samples using alevin-fry to create gene by cell matrices.
The workflow outputs gene expression data in two formats: as SingleCellExperiment objects and as AnnData objects.
Reads from samples are aligned using selective alignment, to an index with transcripts corresponding to spliced cDNA and to intronic regions, denoted by alevin-fry as splici.
These matrices are filtered and additional processing is performed to calculate quality control statistics, create reduced-dimension transformations, assign cell types using both SingleR and CellAssign, and create output reports.
scpca-nf can also process libraries with ADT tags (e.g., CITE-seq), multiplexed libraries (e.g., cell hashing), bulk RNA-seq, and spatial transcriptomics samples.
For more information on the contents of the output files and the processing of all modalities, please see the ScPCA Portal docs.
The default configuration of the scpca-nf workflow is currently set up to process samples as part of the ScPCA portal and requires access to AWS through the Data Lab.
For all other users, scpca-nf can be set up for your computing environment with a few configuration files.
Instructions for using scpca-nf |
|---|
To run scpca-nf on your own samples, you will need to complete the following steps:
- Organize your files so that each folder contains fastq files relevant to a single sequencing run.
- Prepare a run metadata file with one row per library containing all information needed to process your samples.
- Prepare a sample metadata file with one row per sample containing any relevant metadata about each sample (e.g., diagnosis, age, sex, cell line).
- Set up a configuration file, including the definition of a profile, dictating where Nextflow should execute the workflow.
You may also test your configuration file using example data.
For ALSF Data Lab users, please refer to the internal instructions for how to run the workflow on our systems.