Doublet for mixed species

[rjg2186]

Hi @swolock ,

I have a experiment with mixed human and mouse (equal proportion) scRNA-seq. Based on guideline from 10x, the I have provided the expected_doublet_rate as 6.4% expected_doublet_rate=0.064 I am running Scrublet and gets below log. Also, I am adding up the counts for human and mouse based on gene symbol orthologs. Is that fine, or the data should be concatenated for human and mouse ?

Preprocessing...
Simulating doublets...
Embedding transcriptomes using PCA...
Calculating doublet scores...
Automatically set threshold at doublet score = 0.20
Detected doublet rate = 5.1%
Estimated detectable doublet fraction = 45.7%
Overall doublet rate:
	Expected   = 6.4%
	Estimated  = 11.2%
Elapsed time: 13.7 seconds

I am not clear how to identify the final doublet rate for the experiment. Should it be "Detected doublet rate" or "Estimated" For mixed species experiment, is there any specific input required as parameters Please suggest.