nf-core pipeline ·
nf-core/panoramaseq
A Nextflow pipeline for processing sequencing-based spatial transcriptomics data from in-situ arrays.
What nf-core/panoramaseq is
nf-core/panoramaseq is a bioinformatics pipeline designed to process sequencing data derived from in-situ array spatial transcriptomics experiments. Developed by Franco Poma-Soto with backing from Ghent University and the Center for Medical Genetics of Ghent (CMGG), the pipeline automates the end-to-end transformation of paired-end sequencing reads into spatially referenced gene expression matrices.
Spatial transcriptomics workflows require demultiplexing positional barcodes, tracking unique molecular identifiers (UMIs), aligning cDNA reads against reference genomes, and assigning reads to transcript models. Handling these stages manually or with fragmented scripts introduces reproducibility challenges and high computational overheads, particularly during spatial barcode matching.
The pipeline resolves these operational bottlenecks by pairing GPU-accelerated barcode identification with established alignment, deduplication, and quality control utilities. It outputs analysis-ready AnnData H5AD files compatible with standard downstream tools such as Scanpy and Seurat, alongside multi-stage quality reports.
What the pipeline does
The workflow ingests paired-end FASTQ files along with a coordinate-annotated barcode CSV and a samplesheet. Initial quality metrics are collected using FastQC, with an option to downsample reads via seqtk for test runs. Spatial barcodes are called via the GPU-accelerated QUIK tool, followed by UMI sequence extraction using UMI-tools. Cutadapt then removes adapter contamination and low-quality bases before STAR aligns the reads against a user-supplied reference genome or an automated build generated from FASTA and GTF files.
Aligned reads are assigned to genomic features using featureCounts, after which UMI-tools deduplicates and counts UMIs per gene. Count matrices and spatial coordinates are compiled into AnnData H5AD files, which undergo automated structural validation. FastQC, alignment, and quantification logs are ultimately aggregated by MultiQC into a single interactive HTML report.
Key facts
- Pipeline engine
- Nextflow (nf-core)
- Lead developer
- Franco Poma-Soto
- Primary input files
- Paired-end FASTQ (.fastq.gz), samplesheet CSV, spatial barcode CSV
- Primary output files
- AnnData H5AD count matrices, MultiQC HTML report, STAR BAM alignments
- Barcode calling software
- QUIK (GPU-accelerated)
- Read alignment software
- STAR
- UMI handling software
- UMI-tools
- Quantification software
- featureCounts (Subread)
- Required execution environment
- Singularity/Apptainer with NVIDIA GPU and CUDA 12.6 or higher
What you can use it for
- In-situ array expression profiling
- Processing raw paired-end spatial transcriptomics data into unified count matrices for spatial tissue analysis.
- High-throughput barcode identification
- Accelerating spatial barcode deconvolution on NVIDIA GPU hardware for arrays with tens of thousands of spatial features.
- Multi-sample spatial transcriptomics aggregation
- Processing multiple sequencing libraries simultaneously and merging them into a single harmonised H5AD dataset.
Run it here
The research agent runs nf-core/panoramaseq 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
- Delivers rapid barcode calling of approximately 0.75 milliseconds per read on NVIDIA Tesla V100 GPUs using QUIK.
- Produces validated AnnData H5AD files directly compatible with Python and R spatial analysis environments like Scanpy and Seurat.
- Provides built-in index generation to build STAR reference indices directly from genome FASTA and GTF files if pre-built indices are unavailable.
- Incorporates comprehensive quality tracking across raw reads, trimmed sequences, alignments, and deduplicated features via MultiQC.
Watch out for
- Requires an NVIDIA GPU with CUDA 12.6 or higher drivers and Singularity or Apptainer to execute the barcode calling step.
- Restricts barcode length to 36 base pairs and rejection threshold to 8 by default, requiring a container rebuild to change compile-time parameters.
- Requires manual addition of GPU scheduling directives to nextflow.config to integrate with institutional HPC schedulers.
Questions
Can I run nf-core/panoramaseq without my own cluster?
The research agent runs nf-core/panoramaseq 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/panoramaseq under?
MIT, like every nf-core pipeline.
Can the pipeline run entirely on CPUs?
No, the pipeline relies on QUIK for spatial barcode identification, which requires an NVIDIA GPU and Singularity or Apptainer.
What formats are accepted for reference genomes?
The pipeline can take a pre-computed STAR genome index directory with a GTF file, raw FASTA and GTF files to generate an index automatically, or an iGenomes reference identifier.
How are UMI sequences handled?
UMI-tools extracts UMIs using a configurable pattern that defaults to 9 base pairs ('NNNNNNNNN') and performs downstream deduplication per gene.
Can multiple spatial libraries be merged automatically?
Yes, setting the mergecounts parameter to true aggregates all processed samples into a single combined H5AD file.
How can the workflow be tested on smaller data volumes?
The seqtk subsampling parameter sample_size can be set to downsample raw FASTQ reads to a specified count prior to processing.
How to cite it
Philip Ewels, Alexander Peltzer, Sven Fillinger, Harshil Patel, Johannes Alneberg, Andreas Wilm, Maxime Ulysse Garcia, Paolo Di Tommaso & Sven Nahnsen. The nf-core framework for community-curated bioinformatics pipelines. Nat Biotechnol. 2020 Feb 13. doi: 10.1038/s41587-020-0439-x.
Written from nf-core/panoramaseq's own README. Statements citing a figure that does not appear there are removed before publication. Licence and running details come from our registry.