sc/snATAC-seq barcode correction

Hi Cutadapt team,

This is not a standard "issue" but an inquiry regarding sn/scATAC-seq data preprocessing (10x Genomics multiome). I'm using Cutadapt to trim snATAC-seq data reads before alignment with bwa-mem2, and it certainly works admirably, and the parallelization is nicely implemented. My concern relates to barcode assessment and correction.

In the workflow I've been testing barcode correction is done in a separate preprocessing step. The three .fastq files (R1 R2 and I1) are read into an executable and 1) the sequenced barcode is extracted from the correct position in the data for each read, 2) compared to a whitelist of anticipated barcodes (from 10x), and 3) the data are written out with the corrected/mapped barcodes as tags in the read names. Those output fastq files are passed to Cutadapt and ultimately the corrected barcode tags on the read names are used in the downstream analysis after alignment.

Such a barcode correction step appears distinct from the demultiplexing options described in the online documentation, and I am inferring this is not currently an anticipated usage for Cutadapt. Is that accurate? Nevertheless, it may be surprisingly easy to implement and a nice extension to be able to fully preprocess sn/scATAC-seq data in one step. Is this a functionality you are already implementing, or open to including? On what time scale? Thanks in advance!

Best regards,

Sara Knaack

The issue names no source files or tests. Start by reviewing Cutadapt's demultiplexing documentation and comparing it with the described R1, R2, and I1 barcode-correction workflow using 10x Genomics whitelist barcodes before bwa-mem2 alignment. Done would require a defined scope and an agreed implementation plan for integrated barcode extraction and correction.