nf-core pipeline · archived · rnaseq, single-cell, single-nuclei
nf-core/scflow
Analysis pipeline for single-cell and single-nuclei RNA-sequencing data, deprecated in favour of scdownstream.
What nf-core/scflow is
nf-core/scflow is a Nextflow bioinformatics pipeline designed for scalable case and control analyses of single-cell and single-nuclei RNA-sequencing data. It was originally developed by Dr Combiz Khozoie for use at the UK Dementia Research Institute and the Department of Brain Sciences at Imperial College London, with additional contributions from Dr Nurun Fancy and Dr Mahdi M. Marjaneh.
The pipeline was created to take post-demultiplexed sparse gene-cell count matrices through end-to-end processing, spanning quality control, dataset integration, clustering, cell-type annotation, differential expression, and pathway analysis. It structures these steps into a portable workflow that outputs interactive reports alongside standard analysis objects.
This pipeline was deprecated on 13 June 2024 because it did not meet the quality standards of nf-core. Users are directed to use the nf-core/scdownstream pipeline instead.
What the pipeline does
The pipeline takes sparse gene-cell count matrices and runs sample-level quality control, including ambient RNA profiling, thresholding, and doublet or multiplet identification, followed by merged quality control to detect inter-sample outliers. It then performs dataset integration with evaluation metrics, dimensionality reduction using UMAP, t-SNE, or both, and community detection clustering using the Leiden or Louvain algorithms.
Downstream analytical steps include automated cell-type annotation with marker gene characterisation, differential gene expression across categorical or numerical variables, impacted pathway analysis using multiple methods and databases, and Dirichlet modelling of shifts in cell-type composition. Results are compiled into an annotated SingleCellExperiment object and interactive HTML reports for analytical stages.
Key facts
- Pipeline framework
- Nextflow (>=21.04.0)
- Input format
- Post-demultiplexed sparse gene-cell counts matrices, manifest file, sample sheet, and parameter config file
- Primary output
- SingleCellExperiment (SCE) object and step-specific interactive HTML reports
- Underlying R toolkit
- scFlow
- Original developer
- Dr Combiz Khozoie (UK Dementia Research Institute and Imperial College London)
- Current status
- Archived and deprecated on 13 June 2024
- Replacement pipeline
- nf-core/scdownstream
What you can use it for
- Single-cell and single-nuclei RNA quality control
- Inspect ambient RNA contamination, identify doublets or multiplets, set filtering thresholds, and detect inter-sample outliers across experimental batches.
- Dataset integration and cell clustering
- Integrate multiple count matrices with quantitative performance metrics, calculate UMAP or t-SNE embeddings, and group cells using Leiden or Louvain clustering.
- Automated cell-type annotation and marker discovery
- Label cell populations automatically using marker gene characterisation and extract detailed cell-type quality metrics.
- Case-control differential expression and pathway analysis
- Test for differentially expressed genes against categorical or numerical variables, query pathway databases, and model changes in cell composition using Dirichlet distributions.
Strengths and limits
Good at
- Executes full case-control analyses from sparse matrices to annotated SingleCellExperiment objects within a single pipeline run.
- Supports pipeline parameter tuning with execution caching, allowing interrupted runs to restart only downstream steps.
- Generates interactive HTML reports across analytical steps to visualise quality control, clustering, and integration results.
- Parallelises computational steps including individual sample quality control, differential expression, pathway analysis, and Dirichlet composition modelling.
Watch out for
- Deprecated and archived on 13 June 2024 after failing to meet nf-core quality standards.
- No longer maintained, with users advised to migrate to nf-core/scdownstream.
- Requires post-demultiplexed count matrices and cannot process raw FASTQ sequencing files directly.
Questions
Can I run nf-core/scflow without my own cluster?
The research agent runs nf-core/scflow with Nextflow on its own machine — up to 32 cores, 256 GiB of memory and 24 hours per run, which covers the test profile and real single-sample runs. It builds the samplesheet, checks it against the pipeline's schema and reads the results back. When a run is too big for one machine, it gives you the exact command for your own cluster instead.
What licence is nf-core/scflow under?
MIT, like every nf-core pipeline.
Is nf-core/scflow actively maintained?
No. The pipeline was officially deprecated on 13 June 2024 because it did not meet nf-core quality standards, and the repository is archived. The recommended replacement is nf-core/scdownstream.
What inputs are required to run the pipeline?
The pipeline requires three items: a two-column manifest file containing unique sample keys and file paths to sparse gene-cell count matrices, a sample sheet describing the samples, and a Nextflow parameter configuration file.
How can the pipeline outputs be analysed in R?
The pipeline produces a quality-controlled and annotated SingleCellExperiment (SCE) object that can be loaded into R using the read_sce() function from the scFlow R package.
Which clustering and dimensionality reduction methods are implemented?
The pipeline offers community detection clustering via Leiden or Louvain algorithms, and dimensionality reduction through UMAP, t-SNE, or both.
How to cite it
Combiz Khozoie, Nurun Fancy, Mahdi M. Marjaneh, Alan E. Murphy, Paul M. Matthews, Nathan Skene. scFlow: A Scalable and Reproducible Analysis Pipeline for Single-Cell RNA Sequencing Data. bioRxiv 2021 August 19. doi: 10.22541/au.162912533.38489960/v1.
Written from nf-core/scflow's own README. Statements citing a figure that does not appear there are removed before publication. Licence and running details come from our registry.