search

HelenaLC / muscat

Multi-sample multi-group scRNA-seq analysis tools
165 stars 33 forks source link

Problems with RNA ambient #137

Closed hahmu closed 3 months ago

hahmu commented 3 months ago

Dear muscat Team,

Thank you very much for creating such a great package.

I would like to ask if you have any suggestions on dealing with RNA ambient contamination. I didn't sort the cells before sequencing, so I used DoubletFinder to remove the doublets from the data and manually annotated without any issues. However, after using pbDS(), there were many genes that should not be expressed at all.

Any advice or recommendations on how to address this issue would be greatly appreciated.

Thank you in advance for your help.

plger commented 3 months ago

Contamination is a major under-estimated issue, and needs to be addressed independently from doublet calling. This is going to be independent of the muscat workflow, but at the moment the best method for RNA de-contamination is cellbender. It requires a GPU though, so if you don't have that a good alternative is DecontX. You might also want to have a look at the scCDC package/paper: most of the time the problematic contamination is in a handful of genes, and that package aims at identifying those (e.g. so that they can simply be removed from the respective DEA). (BTW for the latter, scDblFinder generally outperforms DoubletFinder and is much faster, though both have an acceptable performance).

hahmu commented 3 months ago

Thank you very much for your fast and helpful response. I appreciate the suggestions you provided on dealing with RNA ambient contamination. I will certainly implement your recommendations and see how they improve my analysis.