Hello, Please find a picture attached of one QC for the 10x analysis report. As you can see the median nr. of genes is 5,597.
I am thinking this is likely due to gross overloading of the sequencer with way too many cells, causing many doublets. But is my assumption correct ?
The other theory is simply that there is very deep sequencing, which causes a lot or basically all of the expressed genes to be amplified. Is this even reasonable?
For referenced other pictures attached, show scatter plots of the nUMI's, nGenes expressed, before and after filtering, and a histogram of the nr. of genes expressed.
Normal doublet removal normally revolves around setting a cap at around 5000-6000 nr. genes expressed and removing them, however if I do this, MOST of my cells are gone and most of the OTHER cells have very low nr. expressed and are likely dead or ruptured cells (<1000 genes expressed).
Do I bite the bullet and take the remaining cells, or is this kind of data considered a failed experiment that needs re-sequencing.
