Single-cell Analysis / R package
dreamlet
Compare expression by cell type across cohorts, including designs with repeated samples.
Best fit
Analysts fitting pseudobulk models that account for donors or other study-level structure.
What you need
R and Bioconductor, counts, cell and sample labels, and a supported design.
What to check
Modeling cannot recover comparisons unsupported by the design. Check coverage and balance within each cell type.