Does dorado demux support duplex calls?

[MostafaYA]

Issue Report

Please describe the issue:

Does dorado demux support the duplex calls. Previously you mentioned this was not implemenetd yet #405?

Thanks

Steps to reproduce the issue:

Please list any steps to reproduce the issue.

Run environment:

• Dorado version: 0.7.2
• Dorado command: dorado duplex -r <model> <seq_dir> > calls.bam && dorado demux calls.bam --sample-sheet <path> --output-dir <dir> --kit-name SQK-NBD114-24 --emit-fastq
• Operating system: Linux
• Hardware (CPUs, Memory, GPUs):
• Source data type (e.g., pod5 or fast5 - please note we always recommend converting to pod5 for optimal basecalling performance):
• Source data location (on device or networked drive - NFS, etc.):
• Details about data (flow cell, kit, read lengths, number of reads, total dataset size in MB/GB/TB):
• Dataset to reproduce, if applicable (small subset of data to share as a pod5 to reproduce the issue):

Logs

• Please provide output trace of dorado (run dorado with -v, or -vv on a small subset)

[malton-ont]

Hi @MostafaYA,

No, dorado duplex does not currently support the --kit-name parameter.

Running dorado demux on duplex called data may work, but will likely have a high proportion of unclassified reads (especially for single-ended barcodes). We have not attempted to evaluate this in any detail.

[mprous1]

I have done it in two steps, dorado duplex first and then demux. There is quite a lot of unclassified reads, but still acceptable level (I think it was certainly less than 30% and maybe less than 20%). I'm surprised that this has not been evaluated before.

Would it be better to do standard basecalling with --kit-name first, then take the classified read IDs and split the pod5 files according to barcode and then do duplex basecalling separately for the samples? It would be rather annoying though and would mean basecalling everything twice. Might be worth maybe if the unclassified reads will be less than 10% in this way.