Adds an R implementation for the HistoCAT neightbourhood analysis (https://www.nature.com/articles/nmeth.4391) that runs from 'CellProfiler output' or 'CellProfiler output'-like data.
The neighbouRhood package has been deprecated. Please install the imcRtools package and use the countInteractions and testInteractions function to perform the neighbourhood analysis.
This require an edge representation of a neightbourhood graph as well as a table that associates each object with a label.
The neightbourhood graph can be exported from object segmentation masks via the Object relationships.csv table.
To do this use the MeasureObjectNeighbors module and add an additional ExportToSpreadsheet module with the option Export all measurement types -> No and select Object Relationships.
Further it needs a table which associates each of the measured oject (with ImageNumber and ObjectNumber) with a label (eg a cluster).
Either clone and install the repository or install devtools to run:
devtools::install_github("BodenmillerGroup/neighbouRhood")
Please follow the vignette https://github.com/BodenmillerGroup/neighbouRhood/blob/master/vignettes/example_permutation_analysis.md
to see an explaination about the concept behind the package and how to use it.
Also all individual functions are documented, ie the documentation can be displayed using: ?neightbouRhood.
In case you need help, please open a Github issue!