In one scRNA-seq sample I encountered the genes: Hbb-bs, Hba-a1 and Hba-a2. These genes appear on top of the list of the highest expressed genes but the 75% percentile of cells have very low counts. Therefore the expression is highly skewed.
Reading from gene cards, they appear to be coding hemoglobin subunits and involved in oxygen transport. I think they may need to be removed before normalization, also because the analysis I'm running does not consider oxygen transport.
Has anyone encountered this situation and what do I do?