Demultiplexing of snAtcSeq data using demuxlet gives an empty output

[yaskermezli]

Hello,

I have a data of snatacSeq of 15 batch of patients (5 patients per batch that are multiplexed ), I need to demultplexe them using genotypes
I am running demuxlet on my single-cell ATAC-seq (snATAC-seq) data, to demultiplexe my batchs

but the output file only contains the header line and no results.

In the log file , I see:
NOTICE [2025/06/28 11:14:06] - Finished processing 0 droplets total
NOTICE [2025/06/28 11:14:06] - Finished writing output files

and in the output.best i have only the header :
BARCODE RD.TOTL RD.PASS RD.UNIQ N.SNP BEST SNG.1ST SNG.LLK1 SNG.2ND SNG.LLK2 SNG.LLK0 DBL.1ST DBL.2ND ALPHA LLK12 LLK1 LLK2 LLK10 LLK20 LLK00 PRB.DBL PRB.SNG1

My input BAM and VCF files seem to be correctly formatted.

There is some flags differences between RNA and ATAC bams but le CB is here, the GT is also present in my vcfs (I have one vcf bu batch)

Are there any specific requirements or known issues when using demuxlet with snATAC-seq data?

Could you advise on possible reasons why no droplets are being processed or written to the output?

Thank you for your help!

​

[yaskermezli]

Hello,

I'm still looking for help / suggestions ?

My be I have to change (somehow) the bam's format ?

Best;

[yimmieg]

Thanks for the message. I’m cc’ing a few of my trainees who have had more experience demultiplexing atac-seq data.

…

On Jul 24, 2025, at 11:07 PM, Yasmina Kermezli ***@***.***> wrote: yaskermezli left a comment (statgen/demuxlet#115) <#115 (comment)> Hello, I'm still looking for help / suggestions ? My be I have to change (somehow) the bam's format ? Best; — Reply to this email directly, view it on GitHub <#115 (comment)>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/ABCQY7ZC2PFI6SNAMOS2XAL3KHCQTAVCNFSM6AAAAACAK2DWNOVHI2DSMVQWIX3LMV43OSLTON2WKQ3PNVWWK3TUHMZTCMJWGUZDENZVGI>. You are receiving this because you are subscribed to this thread.

[marcalva]

Can you try setting --tag-UMI "" to set the BAM UMI field blank? Without knowing much more, can you also verify the chromosome IDs in the VCF and BAM match by string, such that you have "chr1" and "chr1", and NOT "chr1" and "1"?

[yaskermezli]

I'm not using the --tag-UMI argument.
I call demuxlet as follows:

demuxlet
--sam "$BAM"
--vcf "$VCF"
--field GT
--alpha 0.01
--group-list "$OUT/barcodes.txt"
--out "$OUT/${BATCH}_demuxlet"
> "$OUT/log.txt" 2>&1

echo "Results written to: $OUT/${BATCH}_demuxlet.best"
echo

I'm sure that the chromosome names are the same in both the BAM and VCF files.