Quantify ARC-v1 data

[ChristopherBarrington]

I would like to quantify at least the RNA data generated from the ARC-v1 kit (ideally both ATAC and RNA). I have used kb to quantify scRNA-seq before but not data from single nuclei.

Firstly I need to build an index. Since the kit is single nucleus I would need to include the --workflow nucleus argument? The other arguments' default values seem suitable.

When quantifying using kb count I need to again specify --workflow nucleus but I am a little stuck with the -x argument. Looking at the available technologies (kb --list) there is not an entry for ARC-v1 but I can't find the equivalent information for the ARC-v1 kit (or if it would be ok to use 10XV3 for example?).

Any advice on quantifying ARC-v1 data with kb would be greatly appreciated.

[Yenaled]

For -x, you'll need to know where the barcodes, UMIs, and actual biological reads are in your pair of fastq files.

E.g. 10XV3 is equivalent to -x 0,0,16:0,16,28:1,0,0 (a 16-bp long barcode from 1st to 16th base of R1, a 12-bp UMI from 17th to 28th base of R1, and the entire transcriptomic sequences in whole R2).

In other words, -x is formatted as barcode:umi:transcript where each of those three values has three numbers separated by comma (the file, the start position within the read, the end position within the read; all zero-indexed). Note that having a start position of 0 and an end position of 0 means just consider the entire read.

I'm not familiar with ARC-v1, but that information should be specified in the kit somewhere.