bcl2fastq not recognizing multiple barcode-mismatch argument

Question

[Derrik]

I'm calling bcl2fastq with barcode-mismatches set with a series of comma delimited values, which per their documentation is accepted. I checked that the numbers correspond to the actual indices that need lower allowed mismatches, yet the program is still throwing barcode collision errors. If I run the command with just '0' it will succeed, but I'd like to take advantage of the multiple inputs if possible. Am I missing something in the syntax to make it work? I've tried running it with and without spaces after commas, surrounding it in quotes, etc. with no success.

bcl2fastq \
  --output-dir Unaligned_test1 \
  --sample-sheet <...>demux.csv \
  --runfolder-dir <runfolder> \
  --barcode-mismatches 1,1,0,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,0,1,1,1,1,0,1,1,0,0,1,0,1,1,1,0,1,1,0,1,1,1,1,1,1,1,1,1,1,0,1,1,1,1,1,1,0,1,0,1,0,1,1,0,1,1,0,0,1,1,0,1,0,1,1,1,1,1,1,1,1,1,1,0,1,1,1,1,1,1,1,1,1,1,1,0,1,1,1,1,1,0,1,1,0,1,1,1,1,0,1,1,1,1,1,0,1,0,0,0,1,1,0,1,1,1,0,0,0,0

documentation for bcl2fastq

  --barcode-mismatches arg (=1)                   number of allowed mismatches per index
                                                  Multiple, comma delimited, entries allowed. Each entry is applied to
                                                  the corresponding index; last entry applies to all remaining 
                                                  indices.
                                                  Accepted values: 0, 1, 2.

Answer 1

[GenoMax]

yet the program is still throwing barcode collision errors

That makes sense since the run is still considering all indexes at the same time so if you allow one error in a specific index and now it marches another entry then there is collision and bcl2fastq can't bin that read.

We have always done separate runs allowing for perfect matches or 1 error. If you want to use a different setting for some indexes then create subsets of samples in different samplesheets and run the demux multiple times.

[Derrik]

I had a program that calculates the collision of all barcodes in the samplesheet and decides what the maximum allowed mismatches would be, so the values should be 'correct', but I think the argument is not doing what I anticipated, per swbarnes' comment. I think you're right that we'd have to do two separate runs. Thanks!

Answer 2

[swbarnes2]

Here's what my version says about input to barcode-mismatch

--barcode-mismatches arg (=1) number of allowed mismatches per index Multiple, comma delimited, entries allowed. Each entry is applied to the corresponding index; last entry applies to all remaining indices. Accepted values: 0, 1, 2.

I think this lets you control the number of mismatches allowed in the first index versus the second index, not to make some sample indices more flexible than others.

[Derrik]

Ok, that would make sense then.

I somewhat confirmed this by trying configurations of 0,1 and 1,0 as the argument. The 1,0 argument failed, but the 0,1 argument succeeded, which seems consistent with this interpretation.

Thanks!