demultiplexing?

[msport469 (Mike Sportiello)]

Hi,

I have two fastq.gz files that are a forward “raw read” and a reverse “raw read.” I also have the barcode tsv.

How do I import the fastq files and demultiplex everything? I can manifest them in but that doesn’t demultiplex it since I can’t include the metadata in that way, right? Any help would be appreciated.

[Nicholas_Bokulich]

Hi @msport469,
Multiplexed reads with a separate barcodes fastq file can be imported and demultiplexed using the appropriate EMP format and qiime demux emp-paired — see the importing tutorial at qiime2.org for more details.

Good luck!

[msport469 (Mike Sportiello)]

thanks, but i have the barcode tsv not the barcode fastq. Is there a way around this?

[Nicholas_Bokulich]

as in the tsv just maps barcodes to samples but you do not have a separate barcode sequence file?

choose your own adventure:

Where are your barcodes?

1. in-line with the sequences
   a. see the q2-cutadapt tutorial
2. in the header line or some place else
   a. QIIME 2 does not have a way to handle this right now (or maybe ever? this is an unusual format)
   b. figure out a way to extract barcodes from the headers into a separate file (qiime1 had a way to do this I think)
   c. import and process as EMP.

[Luke_Thompson]

Hi @msport469 and @Nicholas_Bokulich,

I’m working with some data that was similarly processed. After some searching, I found on Biostars a package BBMap, available from bioconda, that comes with a script demuxbyname.sh that will do what we want.

If your fastq header looks like this (with barcode or sample number “CAGATC”):

@A00769:25:HHMVKDRXX:2:2101:6162:1047 1:N:0:CAGATC

This command should give you what you need:

demuxbyname.sh in=R1_001.fastq.gz in2=R2_001.fastq.gz out=%_#.fq.gz delimiter=: prefixmode=f