Demultiplexing dual indexed run?


Hi folks -

Is there a good way to demultiplex a dual indexed read run in mothur? We got data back from our sequencing center that hadn’t been demuxed and had the index data in the headers. Like so (CGAGAGTT+CGTCGCTA) - see below. I know this should be pretty easy for them to re-do at the sequencing center but now I’m curious…


@M05104:9:000000000-BFL43:1:1101:15750:1332 1:N:0:CGAGAGTT+CGTCGCTA
@M05104:9:000000000-BFL43:1:1101:15805:1334 1:N:0:TCTTTCTC+TCTTTCCC


mothur does do demultiplexing in make.contigs if you supply an oligos file. however, it sounds like it might be easiest to write a script to do it from the headers.



I am encountering the same confusion here. However, the indexes are not in the headers. I have to do it manually, or with the help of mothur.
The sequencing strategy we used is a bit different – dual indexed paired-end reads with heterogeneity spacers. We used a heterogeneity spacer system, that is ranging from 0 to 7 bp. Should I list out all the spacer in pairs in the oligo file? Or I do not need to list the spacers at all? I believe that the primers are already removed.


You would need to add spacer sequences to your barcodes or the oligos file. make.contigs expects the sequences to start with those sequences.


Hi Pat,

Thanks for your reply. The command runs fine if I do not include the spacer, when I do include the spacer, the following error message popped up:

[ERROR]: cannot mix paired primers and barcodes with non paired or linkers and spacers, quitting.

Making contigs…

So my oligo files have paired barcodes, but the spacers are not paired. Would that be the problem?

When you say, I can add spacer to barcodes, do you mean I just append them?




Eddi - That is correct.