Using Isoquant for single-cell data and de-duplications

[Kyung-TaeLee]

Hi, thank you for providing such a wonderful tool and keeping constant updates.

I am currently trying to use Isoquant for single-cell Nanopore data (10X cDNA library -> genomic ligation sequencing).
I ran Flexiplex for barcode and UMI identification, minimap2 for alignment, and successfully ran Isoquant for quantification. I tried to do de-duplication myself using "OUT.read_assignments.tsv.gz" file but It is not clear to me which reads are actually used for quantification because even if read is assigned to transcript, there are many classification such as unique, inconsistent, etc.
Could you tell me which reads with certain classification are actually quantified for specific feature?

And you mentioned in other post that you have personal de-duplication code that you could share. Could you also share that script by e-mail?

Thank you once again and have a great day.

Best regards,
KyungTae Lee

[andrewprzh]

Dear @Kyung-TaeLee

Thank you for the feedback!

Yes, I work on our own barcode calling and UMI deduplication.
In brief, it is currently in sc_v3 branch. It can be run the same way as IsoQuant with couple additional parameters, such as

Single-cell/spatial-related options::
  --mode {bulk,tenX,double}, -m {bulk,tenX,double}
                        IsoQuant modes: bulk, tenX, double; default:bulk
  --barcode_whitelist BARCODE_WHITELIST
                        file with barcode whitelist for barcode calling
  --barcoded_reads BARCODED_READS [BARCODED_READS ...]
                        file with barcoded reads; barcodes will be called automatically if not provided

Feel free to email me for further details - the pipeline is still in progress, but I already used it in several projects.

Best
Andrey