nf-core pipeline · 10xgenomics, long-read-sequencing, nanopore
nf-core/scnanoseq
A Nextflow pipeline for processing 10X Genomics single-cell and single-nuclei Oxford Nanopore RNA sequencing data.
What nf-core/scnanoseq is
nf-core/scnanoseq is a bioinformatics analysis pipeline built for single-cell and single-nuclei RNA sequencing data generated using 10X Genomics protocols and sequenced on Oxford Nanopore Technologies platforms. It was originally developed by Austyn Trull and Dr Lara Ianov to provide reproducible, containerised workflows for long-read single-cell transcriptomics.
Long-read sequencing captures full-length cDNA molecules, enabling researchers to investigate transcript isoforms and gene expression at single-cell resolution. The workflow is designed around Oxford Nanopore Q20+ chemistry, such as R10.4 flow cells with quality scores exceeding Q20, allowing cell barcode detection and transcript quantification without relying on matched Illumina short-read sequencing data. It can also process data generated with older Nanopore chemistries.
The workflow performs end-to-end processing starting from raw FASTQ files, executing quality control, barcode identification, error correction, genome and transcriptome alignment, deduplication, and quantification. It outputs standard feature-barcode expression matrices that can be imported directly into downstream analytical frameworks such as Seurat.
What the pipeline does
Input FASTQ files are first assessed for quality using FastQC, NanoPlot, NanoComp, and ToulligQC, and can be optionally divided into chunks using split and pigz to accelerate downstream steps. Chopper trims and filters reads prior to barcode identification using BLAZE, which matches cell barcodes against standard 10X or custom whitelists. FASTQ records are then parsed to extract barcodes and unique molecular identifiers (UMIs) and corrected for errors via custom Python scripts, followed by post-extraction quality control.
Reads are aligned to a reference genome, transcriptome, or both using minimap2, and filtered with SAMtools alongside alignment evaluations using NanoComp and RSeQC. Mapped reads receive barcode and UMI quality tags before deduplication with either UMI-tools or Picard MarkDuplicates. Gene-level and transcript-level feature-barcode matrices are quantified using IsoQuant, transcript-level matrices can be generated using oarfish, and initial single-cell quality control is evaluated with Seurat before MultiQC compiles an aggregated quality report.
Key facts
- Pipeline framework
- Nextflow DSL2
- Primary developers
- Austyn Trull and Dr. Lara Ianov
- Supported sequencing chemistry
- Oxford Nanopore Technologies Q20+ (R10.4 flow cells) and older chemistries
- Single-cell library format
- 10X Genomics single-cell and single-nuclei RNA-seq
- Barcode detection tool
- BLAZE
- Alignment software
- minimap2
- Deduplication tools
- UMI-tools or Picard MarkDuplicates
- Quantification tools
- IsoQuant and oarfish
- Input formats
- CSV samplesheet and gzipped FASTQ
- Primary outputs
- Feature-barcode matrices (gene- and transcript-level) and MultiQC reports
What you can use it for
- Full-length single-cell transcript isoform quantification
- Quantify transcript-level and gene-level expression matrices from long-read single-cell datasets using IsoQuant or oarfish to assess alternative splicing across single cells.
- Illumina-free 10X Genomics nanopore processing
- Process single-cell or single-nuclei 10X Genomics libraries sequenced with Oxford Nanopore Q20+ chemistry directly without requiring paired short-read Illumina sequencing.
- Single-cell long-read quality assessment
- Evaluate read qualities, barcode matching rates, alignment metrics, and preliminary single-cell feature distributions across raw and processed reads in a unified MultiQC report.
Run it here
The research agent runs nf-core/scnanoseq 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.
Strengths and limits
Good at
- Processes long-read 10X Genomics data independently without requiring companion Illumina sequencing data.
- Produces gene- and transcript-level feature-barcode count matrices compatible with Seurat and standard single-cell pipelines.
- Provides comprehensive multi-stage quality control from raw reads through to post-alignment and expression matrix evaluations.
- Allows parallel processing of barcode extraction and correction by splitting FASTQ files into configurable chunk sizes.
Watch out for
- Processing large PromethION sequencing runs often requires custom configurations to increase default memory and CPU limits.
- UMI deduplication using UMI-tools can require prolonged execution times, necessitating file splitting by chromosome.
- Transcript-level quantification with oarfish requires a transcriptome FASTA file, whereas IsoQuant requires a reference genome FASTA.
Questions
Can I run nf-core/scnanoseq without my own cluster?
The research agent runs nf-core/scnanoseq 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/scnanoseq under?
MIT, like every nf-core pipeline.
Does nf-core/scnanoseq require paired Illumina short-read data?
No. Because the pipeline is optimised for Oxford Nanopore Q20+ chemistry, it detects barcodes and quantifies transcripts directly from nanopore reads without requiring Illumina sequencing data.
What reference files are required to run IsoQuant and oarfish?
IsoQuant requires a reference genome FASTA file and produces both gene- and transcript-level count matrices. In contrast, oarfish requires a transcriptome FASTA file and produces transcript-level matrices only.
How can barcode extraction and correction runtimes be improved?
Runtimes can be reduced by configuring the split_amount parameter (for example, setting it to 500000 lines), which splits input FASTQ files so that barcode extraction and correction run in parallel on smaller subsets.
How are input samples provided to the pipeline?
Inputs are supplied via a CSV samplesheet specifying the sample identifier, path to the gzipped FASTQ file, and expected cell count for each sample. Technical replicates sharing a sample name are automatically merged.
How to cite it
Austyn Trull, Elizabeth A. Worthey, Lara Ianov, scnanoseq: an nf-core pipeline for Oxford Nanopore single-cell RNA-sequencing, Bioinformatics, Volume 41, Issue 9, September 2025, btaf487
Written from nf-core/scnanoseq's own README. Statements citing a figure that does not appear there are removed before publication. Licence and running details come from our registry.