Tutorial for demuxing paired-end reads with barcodes and primers

[Henrik (Henrik Eckermann)]

Thanks for getting back. Is there a tutorial for how to do demultiplex paired fastq files that have barcodes and primers?

[Mehrbod_Estaki]

Hi @Henrik,
If I’ve read your previous posts right, these are dual-indexed, as in barcodes exist on both the F and R reads, right? In that case, the plugin you want to use is cutadapt demultiplex-paired. There isn’t a tutorial for this particular plugin, but it should be easy enough to follow there.The Atacama tutorial does have a demultiplexing step, but it doesn’t use cutadapt and is not dual-indexed. But still might be helpful. Try the plugin and let us know if you run into any specific problems.

[ATG_UC (Alejandro Trujillo Gonzalez)]

I believe this issue was previously highlighted by @LuSanto in a separate post (does q2-cutadapt support dual indexed reads?). @LuSanto highlighted that the current demux paired-end command ignores the reverse barcode, which when one follows a combinatorial indexing approach, has a tendency to only detect unique forward barcodes. @LuSanto solved this issue by making separate metadata files for each unique forward barcode and demuxing his fastq file multiple times.

I currently have the same issue with my files, but I have so many different barcode combinations that prevent me from using @LuSanto approach. @thermokarst suggested there is an update on q2-cutadapt 2019.10 / cutadapt 2.3 which supports UDI strategies, but the forward and reverse barcodes must be all unique.

Good luck!

@ATG_UC