search

TencentAILabHealthcare / spatialID

34 stars 5 forks source link

what is the difference between cell-type transfering for spatial transcriptomics and deconvolution? #10

Open UnkownNames opened 10 months ago

UnkownNames commented 10 months ago

great work, but I some doubt about the problem of celltype-transfering, for example, I can use Tangram to deconvolute the spatial transcriptomics data and get the celltype-label of each spot. In another aspect, I can use spatialID to transfer celltype label of scRNA to spatial transcriptomics data,It seems that I can get similar effects. But spatialID seems that it is not a method of deconvlolution(no comparing method of deconvolution found in paper), but I can't understand the difference between cell-type transfering for spatial transcriptomics and deconvolution, Could you give some suggestions?

SilversH commented 5 months ago

Thanks for your insterests in our work. Cell type transfering works for cell-level resolution spatial transcrpitomics, in which each data represents one single cell. However, deconvolution methods were designed for those in spot-level, in which each data is a set of cells, or a small bulk of tissue.