wheaton5 / souporcell

Clustering scRNAseq by genotypes
MIT License
168 stars 46 forks source link

No doublet output = 0 doublets? #226

Open bpyenson opened 8 months ago

bpyenson commented 8 months ago

Hi,

I've loved using souporcell for my single nucleus RNAseq analysis. Thanks to Haynes and the team for making such a good program.

On one dataset, though, I saw a result in R* seurat analysis of clusters.tsv that did not show any doublet nuclei. There's two explanations for this: 1) that souporcell detected 0 doublets, and simply didn't print them; or 2) that souporcell was unable to calculate doublets. It did distinguish singlets from unassigned, making me think that explanation 1 above is correct. Has anyone had this issue?

Happy to provide output files if helpful.

Thanks

Code and output in R*: clusters=read.table(paste0("clusters.tsv"),header=T) table(clusters$status)

singlet unassigned

1695 320

bpyenson commented 8 months ago

As an update, the same issue occurs for unassigned, where presumably the absence of printing of results means that unassigned nuclei=0. Here is an example:

clusters=read.table(paste0("clusters.tsv"),header=T) table(clusters$status)

doublet singlet 869 5148