Are frequency heatmaps comparable between conditions

[Carolina-Toste]

Hi, thank you so much for such a great package.

I've been trying to use liana to compare cell-cell communication within my dataset of three conditions. And I was wondering how are the frequency of interactions calculated?

If I subset my Seurat into the different conditions and run the liana pipeline for each would the frequency heatmaps be comparable between conditions (even if for example I have a different number of cells in one condition)?

Thank you!

[dbdimitrov]

Hi @Carolina-Toste, apologies for the delayed response. I only noticed the issue now.

The frequency is quite simpe, it just counts the number of ~significant interactions. For comparisons between conditions I would recommend trying out one protocol with Tensor-cell2cell: https://github.com/saezlab/ccc_protocols

https://www.cell.com/cell-reports-methods/fulltext/S2667-2375(24)00089-4