Nextflow pipeline

Introduction

The pipeline is built using Nextflow, a workflow manager to run tasks across multiple compute infrastructures in a very portable manner. It supports conda package manager and singularity / Docker containers making installation easier and results highly reproducible.

Pipeline summary

1. Merge : Make a Maximum Intensity Projection of all selected markers for a approximated cytoplasm.
2. Split : split this projection and nuclear marker into tiles for better managed memory
3. Segmentation : segment this two markers to get a mask of all cells.
4. Stitch : Glue together generated tiles
5. Compute mask (optionnal) : if cellpose is selected, will transform generated flows into real cells masks.
6. Display outlines : Add a new markers in the tiff file to display outlines generated from masks.
7. Pyramidize : Make new images pyramidizable (multiple resolution to be able to display it by chunk)
8. Quantification : From masks and intensity values, compute feature for each cells in the image.

Quick help

nextflow run orion/MIA/ --help
N E X T F L O W  ~  version 23.04.4
Launching `orion/MIA/main.nf` [admiring_blackwell] DSL2 - revision: 3aad0eac48
------------------------------------------------------------------------

                           __  __ ___   _   
                          |  \/  |_ _| /_\  
                          | |\/| || | / _ \ 
                          |_|  |_|___/_/ \_\

                            v0.0.4dev
------------------------------------------------------------------------

------------------------------------------------------------------------
DISCLAIMER

This software is currently under active development and the results 
have been generated with a non stable version.
The reliability, reproducibility and the quality of the results are 
therefore not guaranteed.

/!\ Do not use the results for any kind of projects /!\
------------------------------------------------------------------------



    Usage:

    The typical command for running the pipeline is as follows:

    nextflow run main.nf --images PATH --markers PATH --profile STRING -profile PROFILES

MANDATORY ARGUMENTS:
    --images  PATH                                                                                    Path to input images (must be surrounded with 
                                                                                                      quotes)
    --markers PATH                                                                                    Path to marker (csv format) one per image (with 
                                                                                                      same name)
    --profile STRING [conda, cluster, docker, multiconda, conda, path, multipath, singularity, test]  Configuration profile to use. Can use multiple 
                                                                                                      (comma separated).

OTHER OPTIONS:
    --name   STRING   Name for the pipeline run. If not specified, Nextflow will automatically generate a random mnemonic
    --outDir PATH     The output directory where the results will be saved

=======================================================
Available Profiles
   -profile test                        Run the test dataset
   -profile conda                       Build a new conda environment before running the pipeline. Use `--condaCacheDir` to define the conda cache path
   -profile multiconda                  Build a new conda environment per process before running the pipeline. Use `--condaCacheDir` to define the conda cache path
   -profile path                        Use the installation path defined for all tools. Use `--globalPath` to define the insallation path
   -profile multipath                   Use the installation paths defined for each tool. Use `--globalPath` to define the insallation path
   -profile docker                      Use the Docker images for each process
   -profile singularity                 Use the Singularity images for each process. Use `--singularityPath` to define the insallation path
   -profile cluster                     Run the workflow on the cluster, instead of locally

See usage.md for a more detailed help about options in command line

Quick run

The pipeline can be run on any infrastructure from a list of input files as follows:

Run the pipeline on a test dataset

See the file conf/test.config to set your test image.

nextflow run main.nf -profile test,multiconda

Defining the '-profile'

By default (whithout any profile), Nextflow executes the pipeline locally, expecting that all tools are available from your PATH environment variable.

In addition, several Nextflow profiles are available that allow:

• the use of conda or containers instead of a local installation,
• the submission of the pipeline on a cluster instead of on a local architecture.

The description of each profile is available on the help message (see above).

Here are a few examples to set the profile options:

Run the pipeline locally, using a global environment where all tools are installed (build by conda for instance)

-profile path --globalPath /my/path/to/bioinformatics/tools

Run the pipeline on the cluster, using the Singularity containers

-profile cluster,singularity --singularityImagePath /my/path/to/singularity/containers

Run the pipeline on the cluster, building a new conda environment

-profile cluster,multiconda --condaCacheDir /my/path/to/condaCacheDir

For details about the different profiles available, see Profiles.

Markers.csv

A marker file is a csv file that provides additional details on the marker of the image. Here is a simple example:

cycle,marker_name,segmentation,normalization
0,01_Hoechst,False,
1,AF1,,
2,04_CD31_Argo515,true,45;6500
3,05_CD45_Argo555L,True,789;45047
4,06_CD68_Argo535,True,72;9376
5,07_Vimentin_Argo550,true,3312;39313
6,08_CD4_Argo572,True,3;2882
7,09_FOXP3_Argo584,True,365;10121
8,10_CD8a_Argo602,True,5749;52408

...

Cycle indicate the laser number used by orion technology (not used), marker name is self explanatory, segmentation is used to create a merge channels with cyto or membrane channels that can help segmentation. Finally normalization is optionnal and will be used to normalize each marker before segmentation and quantification. Other columns can be added but will not be used.

Full Documentation

1. Installation
2. Running the pipeline
3. Output and how to interpret the results
4. Troubleshooting

Credits

This pipeline has been written by Maxime CORBÉ

Contacts

For any question, bug or suggestion, please use the issue system or contact the bioinformatics core facility.