Housekeeping genes scRNA seq

Question

[bioinformatics]

Hello,

I am trying to follow this workflow for scRNA seq analysis. https://broadinstitute.github.io/2020_scWorkshop/data-wrangling-scrnaseq.html#filtering-low-quality-cells

I have loaded the data and computed the percentage of mitochondrial genes for each cell

However, the next step is to load a list of housekeeping genes. A file I do not have. Does anyone have any idea on where to locate this file.

Thanks!

Kind regards,

[ATpoint]

You don't need this. It's not standard to do this "housekeepers per cell", and generally the term "housekeeper" is poorly defined. Just do the normal "% mitochondrial reads", "total counts per cell" and "detected genes per cell", and then filter cells that are a certain cutoff away, like 3xMAD. If you have multiple celltypes, batches, or conditions, then do QC per batch/celltype/condition. As mentioned already, go easy on cutoffs, rather be lenient than too strict, as you can always go back and make it stricter if you realize downstream that you might still have poor-quality cells in there that cause problems.

[fracarb8]

a simply google search returns the file you are looking for.

Answer 1

[shiyang_bio]

Hi Ram, From my experience, filtering cells based on the expression of housekeeping genes is not a very common step. You can take a look at those single cell papers, most of them only use UMI count, gene count and MT percent. You can have a try, there is no downside. But do not filter too hard. Be soft. Do not remove too many cells. Finally you will see that some cells with low UMI still value.

Answer 2

[Radu Tanasa]

I agree with ATpoint and shiyang_bio that filtering housekeeping genes is not something you should normally do. I have seen people filter based on ribosomal genes in addition to mitochondrial, but that too is not standard practice. Here's a list of genes if you'd like to try it out: http://software.broadinstitute.org/gsea/msigdb/download_geneset.jsp?geneSetName=KEGG_RIBOSOME&fileType=txt