Once you have chosen a study and Explorer Version, next define which features you wish to view in the plots.
Cell type: How much of each of the defined cell populations are in the regions in each cohort? These may have been defined in the gating extension, or in analyses in workbench.
Cell interaction: Which cell types interact more frequently in one cohort vs. another? For this analysis, interaction is defined as contact between the voronoi expansions from each cell’s centroid. This is not normalized for the frequency of each of the cell types, so it can be driven by both interactions and cell frequency.
Cell interaction (log2 odds): An odds ratio, how much more likely are you to see any given pairwise interaction between cell types than you would expect by chance? This odds ratio is then log2 transformed. We compute this feature as described in Goltsev et al., 2018.
Neighborhoods: How much of each of the regions in the cohorts belongs to each of the defined Cell Neighborhoods?
Next:
Return to Home Directory