Cell phenotyping using CELESTA tailored to MCMICRO outputs, formatted for implementation in nf-core modules

mcmicro nf-core module for cell phenotyping using CELESTA

CELESTA_CLI.R

CELESTA script with a CLI implementation using the optparse R Package.

Overview of changes compared to the original Script

• Added the optparse R Package
• implemented CLI input options as a list
• Checking if mandatory inputs were provided
• Changing columns of image_data provided to fit requirements for CELESTA
• If optional vector inputs are not given, a custom vector is created that consists of only 1s
• Checking if folder output and a title were provided
• Surpressing the default save_result output from AssignCells()
• Creating the desired CSV output by getting final_cell_type_assignment from the CelestaObj and binding it to the input from `--image_data'
• Creating a quality inspection .csv file from marker_exp_prob of the CelestaObj.

Usage

Option	Description	Mandatory
-i/--image_data	Path to the quantification output as a .csv (e.g., after running MCMICRO). Includes marker intensities, CellID, and X/Y columns.	Yes
-s/--signature	Path to the signature matrix for cell type definition (prior marker info) as a .csv file. Description can be found here.	Yes
--high	Path to a .csv file with 2 rows defining high thresholds for anchor cell (row 1) and index cells (row2).	Yes
--low	Path to a .csv file with 2 rows defining low thresholds for anchor cell (row 1) and index cells (row2).	No
-o/--output	User-defined path to a output file, if not provided the result files will be saved in the current workingdirectory	No
-t/--title	User-defined tag used to define the project title inside the CELESTA algorithm and provided in the result .csv file.	No

Docker Usage

To build the container:

git clone https://github.com/SchapiroLabor/mcmicro-celesta.git
docker build -t mcmicro_celesta:latest .

To run the docker container:

## Note: Mount the local volume to the local folder in the directory
docker run --platform linux/amd64 -it --rm -v $(pwd):/local celesta Rscript /local/CELESTA_CLI.R <input_parameters>

To pull the container from the Github container registry (ghcr.io):

## Login to ghcr.io
docker login ghcr.io

## Pull container
docker pull ghcr.io/schapirolabor/mcmicro-celesta:v0.0.2