What this tab does: Tests which genes are up or down regulated between two conditions (e.g. treated vs control) within a specific cell type.

Since single cells aren't true independent replicates, we first pick one cell type (e.g. Epithelial Cells), then group all cells of that type from the same biological sample together by adding up their counts — this is called pseudobulk. Each sample becomes one data point, just like a traditional bulk RNA-seq experiment. Then we use standard statistical methods (voom + limma) to find genes that are significantly different between your two conditions.

How to use it:

1. Pick a cell type — you're asking: in this cell type, what genes change between conditions?
2. Pick your replicate column — this is your sample or patient ID (not cluster)
3. Pick your condition column — this is what you're comparing (e.g. treatment, disease)
4. If more than 2 conditions — select which 2 groups to compare
5. Check the tables on the right — make sure each sample maps to only one condition
6. Click Run — results appear as volcano/MA plots and a downloadable table