SCA-IRCM / SingleCellSignalR_v1

R package
26 stars 17 forks source link

How to construct ad hoc benchmark #16

Open eleozzr opened 4 years ago

eleozzr commented 4 years ago

Hi SCA-IRCM team,

In your paper (SingleCellSignalR), you said: We thus designed an evaluation where candidate LRdb pairs, restricted to ligands, receptors, and cell types covered in Tref, could be considered correct if both the ligand and the receptor TPM were above 10. But there are 144 different cell types in FANTOM5 database and no exactly named by "Macrophage" or T cells. My question is how you rename the original cell type label, for example, T cells include CD4+ T cells and CD8+ T cells, did you collectively call them T cells. Could you provide me some detail?

For the second benchmark dataset, I didn't found raw proteomics count data. I am considering that could you provide me where I can access this data? Would be very appreciative.