nf-core pipeline · molecular-pixelation, pixelator, pixelgen-technologies
nf-core/pixelator
Processes Proximity Network Assay sequencing data into single-cell protein abundance and interactomics graphs.
What nf-core/pixelator is
nf-core/pixelator is a bioinformatics pipeline engineered to process raw sequencing reads generated by the Proximity Network Assay (PNA). Originally created for Pixelgen Technologies AB by Florian De Temmerman, Johan Dahlberg, and Alvaro Martinez Barrio, the workflow standardises the analysis of spatial proteomics data at single-cell resolution.
The pipeline takes experimental design metadata and paired-end sequencing reads via a samplesheet, managing quality checks, molecular error correction, graph construction, and spatial characterisation. It yields single-cell protein abundance measurements, interactome networks stored in PXL files, and graphical summaries for downstream exploration.
What the pipeline does
The workflow begins by running quality control checks on input sequencing reads and assembling amplicons using the pixelator single-cell-pna amplicon module. Amplicons are subsequently demultiplexed into groups according to their marker assignments. Next, the pipeline performs error correction and deduplication to collapse amplicons into counts of original molecules, which form the edge list of an interaction graph. Components of the graph are resolved to define putative individual cells, and in hashed multiplexed experiments, sample calling assigns these components back to their respective samples.
Following initial cell graph assembly, the pipeline applies graph denoising algorithms and performs spatial analysis on the single-cell networks. It then computes three-dimensional graph layouts to assist with visualising cell topologies and structures. Finally, an experiment report is produced via PixelatorES to summarise the assay results.
Key facts
- Underlying software
- pixelator, PixelatorES
- Assay compatibility
- Proximity Network Assay (Proxiome kit v1 and v2)
- Input formats
- CSV (samplesheet), FASTQ (paired-end .fastq.gz)
- Output formats
- PXL, Proxiome Experiment Summary
- Latest release
- 5.0.1
- Pipeline authors
- Florian De Temmerman, Johan Dahlberg, Alvaro Martinez Barrio
What you can use it for
- Single-cell protein interactomics
- Derive protein-protein interaction networks and spatial proximity topologies for single cells from raw sequencing data.
- Single-cell protein quantification
- Quantify panel-specific protein abundances across individual cells following amplicon error correction and molecule collapsing.
- Multiplexed PNA sample deconvolution
- Demultiplex and call individual samples from pooled, hashed sequencing experiments using Proxiome kit v2 assays.
- 3D cellular graph visualisation
- Generate three-dimensional coordinates and graph layouts representing the spatial organisation of proteins on individual cells.
Run it here
The research agent runs nf-core/pixelator 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
- Carries out complete end-to-end processing from raw FASTQ quality control to 3D graph layout generation.
- Supports both unhashed single-sample assays (Proxiome kit v1) and hashed pooled libraries (Proxiome kit v2).
- Uses graph component derivation and denoising algorithms to reconstruct individual putative cells from molecular edge lists.
Watch out for
- Does not support Conda or Mamba environments due to upstream dependency restrictions, requiring execution inside Docker or Singularity containers.
- Does not support Molecular Pixelation (MPX) assay data in current releases, as MPX support ended after release 2.3.1.
- Requires explicit home directory mounting flags (NXF_APPTAINER_HOME_MOUNT or NXF_SINGULARITY_HOME_MOUNT) when using newer Nextflow versions to avoid dependency failures.
Questions
Can I run nf-core/pixelator without my own cluster?
The research agent runs nf-core/pixelator 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/pixelator under?
MIT, like every nf-core pipeline.
Can this version of the pipeline process Molecular Pixelation (MPX) data?
No. Pipeline versions 4.0 and later only support Proximity Network Assay (PNA) data. Analyses involving Molecular Pixelation must use pipeline release 2.3.1.
Why is the pipeline failing when run with a Conda profile?
Conda and Mamba environments are not supported due to issues with upstream software dependencies. Container runtimes like Docker or Singularity must be used instead.
What information is needed in the input samplesheet?
The samplesheet requires pool IDs, sample identifiers, condition tags, design specifications, panel references, and paths to paired-end FASTQ files, alongside hash indices for hashed runs.
Which assay kits are supported?
The pipeline supports both non-hashed Proximity Network Assay data (Proxiome kit v1) and hashed pooled data (Proxiome kit v2).
How to cite it
Filip Karlsson, Michele Simonetti, Christina Galonska, Max Karlsson, Hanna van Ooijen, Tomasz Kallas, Divya Thiagarajan, Maud Schweitzer, Ludvig Larsson, Vincent van Hoef, Pouria Tajvar, Johan Dahlberg, Florian De Temmerman, Louise Leijonancker, Sylvain Geny, Rikard Forlin, Erika Negrini, Stefan Petkov, Lovisa Franzén, Jessica Bunz, Christine Moge, Henrik Everberg, Petter Brodin, Alvaro Martinez Barrio, Simon Fredriksson. Single-Cell Protein Interactomes by the Proximity Network Assay. bioRxiv 2025.06.19.660329; doi: 10.1101/2025.06.19.660329
Written from nf-core/pixelator's own README. Statements citing a figure that does not appear there are removed before publication. Licence and running details come from our registry.