Implicit requirement in `split` that the column specified with `--demux-rid` must occur before `--demux-code`

[patbohn]

Hi Hasindu, thanks for this super useful tool! I'm currently demultiplexing a large dataset and am supplying custom demultiplexing data.

I noticed that when I have a tsv file where the --demux-code column is occurring before the --demux-id column, I get this error:

[bsum_parsehdr::ERROR] Invalid demux TSV header: missing '[my_column]'

I believe this is due to how the columns are searched. When I swap the columns in my tsv file the splitting works.

I think a more informative error message could help debugging this faster.

[hasindu2008]

Hello

Thanks for reporting this. It is a very good catch. We will fix this after the holidays, ideally to find the column irrespective of the order provide!

@sashajenner please take a note of this.