I basically follow the turtor [Brief tutorial for CellChat analysis of a single spatially resolved transcriptomic dataset] and apply into a visium hd dataset of ctx. The numbers of spots of filtered raw data were too large to perform this turtor (always fail) and thus I just kept singlets from RCTD results. And the values were changed when perform computeCommunProb based on visium hd data. However, the workflow worked fine, but the count (0~20) and weight (0- 3e-6) of net results were too low to perform plot. Does it make sense for HD datasets? How do I improve the results?
cellchat <- computeCommunProb(cellchat, type = "truncatedMean", trim = 0.1,
distance.use = TRUE, interaction.range = 80, scale.distance = 0.12,
contact.dependent = TRUE, contact.range = 16)
I basically follow the turtor [Brief tutorial for CellChat analysis of a single spatially resolved transcriptomic dataset] and apply into a visium hd dataset of ctx. The numbers of spots of filtered raw data were too large to perform this turtor (always fail) and thus I just kept singlets from RCTD results. And the values were changed when perform computeCommunProb based on visium hd data. However, the workflow worked fine, but the count (0~20) and weight (0- 3e-6) of net results were too low to perform plot. Does it make sense for HD datasets? How do I improve the results?
cellchat <- computeCommunProb(cellchat, type = "truncatedMean", trim = 0.1,
distance.use = TRUE, interaction.range = 80, scale.distance = 0.12,
contact.dependent = TRUE, contact.range = 16)