nf-core pipeline · bioformats, image-analysis, image-processing
nf-core/mcmicro
An end-to-end pipeline that transforms multi-channel whole-slide microscopy images into single-cell data.
What nf-core/mcmicro is
nf-core/mcmicro is a Nextflow bioinformatics pipeline designed to process highly-multiplexed imaging data produced by methods such as Cycif, MIBI, CODEX, and SeqIF. It converts multi-channel whole-slide microscopy images into single-cell datasets by performing image registration, cellular segmentation, and single-cell quantification.
The pipeline is an nf-core port based on version 1.0 of MCMICRO, written by Rob Young, Alex Barberá Mourelle, and Jeremy Muhlich from the Laboratory of Systems Pharmacology at Harvard Medical School. The 'MC' stands for 'Multiple Choice', reflecting an architecture designed to support interchangeable software modules across its core processing stages.
Handling multiplexed spatial imaging requires orchestrating complex computer vision and quantification steps across multiple imaging cycles and channels. nf-core/mcmicro standardises this workflow into a reproducible end-to-end process that accepts raw cycle tiles and marker metadata, yielding spatial single-cell feature arrays ready for downstream tissue analysis.
What the pipeline does
The workflow ingests microscopy image files alongside user-provided samplesheet and markersheet CSV files that detail cycle numbers, image paths, and channel marker names. Image cycles first undergo alignment using Ashlar for image registration. The pipeline can also optionally run background and shading correction, background subtraction, and accommodates tissue microarray (TMA) designs.
Following alignment, a pre-segmentation image is generated and processed by segmentation modules, with Cellpose and Mesmer currently supported and able to run in parallel. Finally, segmented cell masks are analysed by MCQuant for quantification, generating a cell-by-feature spreadsheet array.
Key facts
- Pipeline engine
- Nextflow
- Pipeline framework
- nf-core
- Latest release
- 2.0.0
- Input image format
- ome.tiff
- Input metadata formats
- CSV (samplesheet and markersheet)
- Primary outputs
- Pre-segmentation image, segmentation mask image, cell x feature array spreadsheet
- Registration software
- Ashlar
- Segmentation software
- Cellpose, Mesmer
- Quantification software
- MCQuant
What you can use it for
- Multi-cycle Cycif slide processing
- Register cyclic immunofluorescence whole-slide images across iterative staining rounds, segment nuclei and cell bodies, and quantify marker expression into single-cell data tables.
- Multiplexed tissue microarray analysis
- Process whole-slide imaging data generated from tissue microarrays across imaging cycles with integrated registration and feature quantification.
- Parallelised cell segmentation benchmarking
- Run Mesmer and Cellpose segmentation modules simultaneously on the same multi-channel datasets to compare segmentation masks and derived single-cell measurements.
Run it here
The research agent runs nf-core/mcmicro 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
- Offers modular 'Multiple Choice' tool options, enabling multiple segmentation tools like Cellpose and Mesmer to execute in parallel.
- Accommodates iterative staining cycles and multi-channel data natively using standardized sample and marker sheets.
- Includes optional preprocessing capabilities such as shading correction, background subtraction, and support for tissue microarrays.
Watch out for
- The pipeline is under active development as part of an ongoing porting process from the original MCMICRO implementation.
- Pipeline parameters cannot be supplied through custom configuration files using the Nextflow -c option; they require CLI flags or the Nextflow -params-file option.
Questions
Can I run nf-core/mcmicro without my own cluster?
The research agent runs nf-core/mcmicro 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/mcmicro under?
MIT, like every nf-core pipeline.
What imaging platforms are compatible with nf-core/mcmicro?
The pipeline processes highly-multiplexed imaging data from technologies including Cycif, MIBI, CODEX, and SeqIF, using ome.tiff image files.
What files are required as pipeline inputs?
You need a samplesheet CSV defining sample identifiers, cycle numbers, and image tile paths or URLs, along with a markersheet CSV listing channel numbers, corresponding cycle numbers, and marker names.
Which segmentation algorithms are currently available?
The pipeline currently supports Cellpose and DeepCell Mesmer, which can be run in parallel.
What output files does the pipeline produce?
It outputs a pre-segmentation image file, a segmentation mask image, and a single-cell by feature array spreadsheet.
How to cite it
Schapiro et al. 2022 Nat. Methods (https://www.nature.com/articles/s41592-021-01308-y)
Written from nf-core/mcmicro's own README. Statements citing a figure that does not appear there are removed before publication. Licence and running details come from our registry.