Hi folks!
I would be interested to hear your opinions (and recommendations) on how to best normalize mRNA counts of single cells (96 cells, UMI-deduplicated 3' counts) for cell sequencing depth. We are not working on the full transcriptome, but rather use a targeted approach for a specific panel of genes (~30). Ideally, we'd like to avoid ERCC spike-ins. The main driving question would be the comparison of the relative profile of expression of these genes between the cells.
I was considering DESeq2's size factors, or the 'computeSumFactors' function of the scran package. Not sure though if either of these would be appropriate for the small number of genes present in the expression matrix...? Or just a simple 'divide gene count by total library count'...?
Thank you!
I would be interested to hear your opinions (and recommendations) on how to best normalize mRNA counts of single cells (96 cells, UMI-deduplicated 3' counts) for cell sequencing depth. We are not working on the full transcriptome, but rather use a targeted approach for a specific panel of genes (~30). Ideally, we'd like to avoid ERCC spike-ins. The main driving question would be the comparison of the relative profile of expression of these genes between the cells.
I was considering DESeq2's size factors, or the 'computeSumFactors' function of the scran package. Not sure though if either of these would be appropriate for the small number of genes present in the expression matrix...? Or just a simple 'divide gene count by total library count'...?
Thank you!