Hello,
I have 3 samples and each of them contain 4 totalseq B biolegend hashtags from biolegend. The hashtags are used to demultiplex the initial 3 samples into 4 samples. When running scrublet is it correct to run it separately for each of the 3 initial samples? I know the expected doublet rate for the 3 initial samples since I know the number of cells from cell ranger while I don't know the expected double rate for the 4 demultiplexed samples until after I use hashsolo.
Thank you
•
written
by
bioinfo •
1 answer
Yes, running it separately for each of the 3 original samples (feeding this variable to scrublet as batch_key) should be the way to go since these were run in physically separate microfluidic channels (assuming this is 10x or a similar droplet tech).
Ah, okay. Looking at HashSolo, correct me if I'm wrong, it seems like it does demultiplexing and doublet scrubbing at the same time. In that case, you might not even need Scrublet?
If you think you do, maybe you can use it to "pre-scrub" the doublets first on the original sample level, then let HashSolo leverage the cell hashing info to do more. Then it would be a matter of seeing how many cells were flagged/removed at each step and seeing if, combined, it adds up to roughly the percentage of doublets you're expecting (though I probably wouldn't worry too much if it didn't).
Log in to answer this question.
-
Which key should I use with Harmony?
written by bioinfo •
-
Doublet Detection in SMART-Seq2 single-cell data
written by luca8650 •
-
Which method I should use to process cellranger data
written by zizigolu
-
Recommended workflow to pre-process and demultiplex scRNAseq data based on the HTOs and ADTs
written by mete.han.celebi •
-
Does the order of HTOs matter?
written by bioinfo •
-
Do I need to remove doublets after before or after running hashsolo?
written by bioinfo •
-
Tutorial: Single-cell RNA-seq: Preprocessing: Data integration and batch correction
written by Julia Ma •
-
How to choose threshold for scrublet?
written by bioinfo •
-
How to demultiplex data with totalseq B antibodies using hashsolo?
written by bioinfo •
-
5scRNA (GEX, Cell surface, TCR) fastq demultiplexing sub-sequent analysis
written by theodore
Scrublet generates a simulated distribution from the given dataset and computes a threshold for the doublet scores.