Very interesting work! The combined analysis of gene expression and genomic data at single cell resolution is a real strength. In three independent scRNA-seq datasets from HNSCC (including HPV+ HNSCC, HPV-negative HNSCC and normal tonsil epithelium) we (Smith et al EMBO J. 2025 (PMID 39548236)) observed peak A3A expression in differentiating keratinocytes and peak A3B expression in cycling keratinocytes but minimal expression of either gene in resting basal cells. We also provided evidence at the protein level (IHC in tumours) and in cultured keratinocytes (A3A, A3B mRNA expression and A3A RNA editing activity) that A3A is expressed in differentiating, rather than basal cells. From what I can see from your UMAP plots in Fig 1 and in Supplemental Fig 1 in your manuscript, the peak A3A expression and many of cells displaying greatest SBS2 enrichment appear to be 'stratified squamous epithelial cell' (if I am interpreting the key in Supp Fig 1 correctly), which I suspect is equivalent to what we called 'differentiating epithelial cells' (i.e. cells that express genes such as Involucrin and cytokeratin 10, which would not be expected to be present in basal cells). So I think your observations are consistent with ours but that we are using different terminology to describe the relevant epithelial cell populations. I'd be interested in your thoughts on this, and it might be worth looking at the additional scRNA-seq datasets that we analysed in our study to compare them with your dataset.