CellProfiler4 pipelines

Custom CellProfiler4 pipelines for SIMPLI

SIMPLI requires two CellProfiler4 pipelines:

• Image preprocessing pipeline: specified by the cp4_preprocessing_cppipe parameter.
• Cell segmentation pipeline: specified by the cp4_segmentation_cppipe parameter.

SIMPLI compatible CellProfiler4 pipelines

SIMPLI uses CellProfiler4 pipelines in .cppipe format built with CellProfiler4 versions >4.0.7.
To be compatible for use by SIMPLI a CellProfiler4 pipeline needs to be able to be run in a non interactive environment. Input images need to be loaded using the LoadData module instead of the Images, Metadata and NamesAndTypes modules. The pipeline should be designed to process images from one sample at a time, without any grouping. This is because, when possible, parallelization is achieved by running concurrently multiple instances of the pipeline on separate samples.

Additionally, the two pipelines have required output files and naming conventions described in the following sections.

Image preprocessing pipeline

The image preprocessing pipeline is used to produce all the images that will be used as input for:

• Pixel level analysis.
• Cell segmentation.

The output images should be in 16 bit integer format and named following this convention: SAMPLE/SAMPLE-LABEL-Preprocessed.tiff .
Image output in CellProfiler4 is performed by the SaveImages module.

Cell segmentation pipeline

Cell segmentation pipeline is used to identify all cells of interest and perform all single cell measurements. Its output is used as input for the cell level analysis, and it must include:

• Cell Masks:

  For each sample the cells of interest should be converted to an image with the ConvertObjectsToImage module. The resulting image should be output in 16 bit integer format and named following this convention, using the SaveImages module: SAMPLE-Cell_Mask.tiff

• Cell measurements tables:

For each sample a .csv format file with all the required measurements for the cells of interest must be produced and saved as: SAMPLE_NAME-Cells.csv. The required measurements are all the measurements that will be used for:

• Cell type identification
• Cell clustering

These measurements are specified in the cell types metadata configuration file. In CellProfiler4 these measurements are performed with the MeasureObjectIntensity module. These measurements should then be selected and exported with the ExportToSpreadsheet module.

Additional supported CellProfiler4 plugins

Currently SIMPLI supports the following additional CellProfiler4 plugins from the ImcPluginsCP suite:

• cliprange.py
• correctspilloverapply.py
• correctspillovermeasurements.py
• cropimage.py
• exportvarcsv.py
• masktobinstack.py
• measureimageintensitymultichannel.py
• measureobjectintensitymultichannel.py
• saveobjectcrops.py
• smoothmultichannel.py
• stackimages.py
• summarizestack.py
• transformbinary.py