Hello,
Should we use Doublet detection first on the real matrix, then remove the doublets first and then use ambient RNA correction ?
Or should we use DecontX first and take ambient RNA removed matrix as default matrix and use it in the doublet detection tool like scdblFinder ?
•
written
by
carolofharvest •
1 answer
The later, remove the potential RNA present in all droplets then call potential doublets.
I’ve noticed that DecontX can also remove cell type specific marker expression from cells. For example, an NK cell might show low expression of platelet or B cell markers, and DecontX removes those platelet and B cell signals from the NK cell.
But what if that “NK cell” is actually a doublet? In that case, if I correct the expression matrix with DecontX first, wouldn’t doublet detection be less likely to flag it as a doublet, even if it really is one?
Usually if such gene markers expression are mitigated it is due to their general expression across your dataset.
As a sanity check, you can keep your raw matrix and see how much of your B cell markers are been removed from your NK cluster compared to other clusters.
Log in to answer this question.
-
Best practice for combining Ambient RNA correction and sc.pp.regress_out() in scRNA-seq
written by messy_cat •
-
DecontX for ambient RNA removal on 43 samples
written by kato.nauwelaers •
-
Doublet Detection for Intestinal Immune Cells in scRNA-seq
written by s •
-
Do I need to remove doublets after before or after running hashsolo?
written by bioinfo •
-
Ambient RNA removal method that generates whole (integer) counts
written by EK •
-
Forum: Steps after doublet detection in single cell data
written by ishwarvh •
-
Do I need to perform basic filtering before using scrublet?
written by bioinfo •
-
Why runing scDblFinder before and after removing low QC cells gives different results
written by Assa Yeroslaviz
-
ambient RNA removal tool preferences?
written by benjamin.pyenson •
-
Cholmod error when using decontX package
written by abiuma •
