search

jinworks / CellChat

R toolkit for inference, visualization and analysis of cell-cell communication from single-cell and spatially resolved transcriptomics
GNU General Public License v3.0
240 stars 34 forks source link

Visium data #89

Open anjrao17 opened 6 months ago

anjrao17 commented 6 months ago

Hello, thanks for building this great tool and for your answer https://github.com/jinworks/CellChat/issues/83 . Follow up question regarding Visium data, what do you recommend be done for datasets where there are spots with multiple celltypes present as in often the case with the Visium data? Is there a way to give an input such that a spot could be part of multiple cell groups for. eg provide multiple metadata input columns, one per celltype such that a spot could be positive for multiple celltypes? Or a way to weight the analysis based on the spot proportion values per cell type (output of deconvolution methods like RCTD) ?

sqjin commented 6 months ago

@anjrao17 Hello, we thought about this question before. I agree that considering the proportion is particular useful for visium, but do you have any idea on how to estimate the expression level of each cell type within the spot?

pedriniedoardo commented 6 months ago

@sqjin cell2location has a routine to do this. https://cell2location.readthedocs.io/en/latest/notebooks/cell2location_tutorial.html#Estimate-cell-type-specific-expression-of-every-gene-in-the-spatial-data-(needed-for-NCEM).

sqjin commented 6 months ago

Thank you @pedriniedoardo Will take a look to see if it is suitable to use for cell-cell communication inference.