Staining (visualisation)

[caravagn]

We need to implement afunction that colour cells (from the tissue simulation) by their original colour (species) if they have a certain set of mutations given in input. Cells lacking those mutations will be either coloured in gray, or white (optionally).

That is valled virtual statining and can be used to explain clonal/ subclonal mutations and other features.

[albertocasagrande]

Closed in 735ce4d by the plot_tissue() parameters focus_function and alpha_function. See plot_tissue().

[caravagn]

Hi,

https://caravagnalab.github.io/ProCESS/1.3/reference/plot_tissue.html

it does not seem to report the params. Also, I am not sure this is the staining I suggested. The one I meant takes in input a list of mutations, and colours only the cells that have those mutations (shadows the others)

[albertocasagrande]

The tissue simulation is mutation agnostic: it does not deal with cell mutations. Only the cells in the phylogenetic forest are labelled by their mutations. Thus, we cannot implement a function that exclusively takes as the input a tissue simulation and a list of mutations and highlights the cells having these mutations. Even if we admitted as input the phylogenetic forest too, the plot of the last tissue status would not contain cells labelled by their mutations because cells are removed from the tissue simulation as soon as they are sampled.

The plot_tissue()'s parameter focus_function is meant to provide a rapid way to highlight some groups of cells in the plot. If you focus on the phylogenetic forest cells, this parameter can be used to implement something similar what you requested. You cannot set the colors of the grey area, but you can use the parameter color_map to add the grey colors to the legend.

A possible "trick" to obtain what you requested is to (1) sample all the cells in the tissue simulation, (2) build the phylogenetic forest, and (3) plot the tissue as it was before the last sampling focusing on cells with the aimed mutations.

As an alternative to the full tissue sampling, I can implement a TissueSimulation's method is_ancestor_of(id1, id2) that tests whether the cell having identifier id1 is an ancestor of the cell with identifier id2. This method in combination with PhylogeneticForest$get_first_occurrences() could be used to identify the cells that may have those mutations because of inheritance, but it would not help in excluding the cells outside the phylogenetic forest that lost the mutations due to deletions.

[albertocasagrande]

I guess we can close this issue again.