We recently had a demultiplexing fail when the entered barcodes were missing leading 0's, and this caused demultiplexing to silently fail/miss a bunch of barcodes.
We should implement 2 things in response to this:
- If barcodes are mis-formatted (e.g. no leading 0s, still has the leading NB/RLB), should error.
- We should probably check for incorrect barcodes.tsv (e.g., if there's a suspiciously small # of reads matching an included barcode or suspiciously large matching excluded barcode)
We recently had a demultiplexing fail when the entered barcodes were missing leading 0's, and this caused demultiplexing to silently fail/miss a bunch of barcodes.
We should implement 2 things in response to this: