digitalcytometry / ecotyper

EcoTyper is a machine learning framework for large-scale identification of cell states and cellular ecosystems from gene expression data.
Other
177 stars 41 forks source link

Spot-Level Immune Cell Deconvolution in Spatial RNA-seq Data #89

Open sallyseullee-0821 opened 4 months ago

sallyseullee-0821 commented 4 months ago

Hello CIBERSORTx/Ecotyper Team,   I am currently analyzing spatial RNA-seq data that consists of tumor, fibrotic, and normal tissues. I have already clustered the data and annotated each cluster as combinations of cell types that are not immune cell types. My goal is to deconvolute the immune cell composition at the level of individual spots within the dataset.   From my review of the available documentation, it appears that the deconvolution process is primarily designed to operate on a cluster-by-cluster basis.   Could you advise if there is a functionality within CIBERSORTx or Ecotyper that supports spot-level immune cell deconvolution? If not, do you have any recommendations on how I might adapt the current methodologies to achieve this?   Thank you for your assistance in advance!