Demultiplexing data prior to SRA upload - Help!

[coralnerd]

I have a set of 454 amplicon data that comprises six multiplexed (barcoded) samples.

I've already processed and analysed this data through Qiime, but I'm currently trying to get the raw reads uploaded in the NCBI SRA.

I can upload the data as a single .sff file, but it would make more sense to demultiplex it first and upload separate fatsq files for each sample.

I tried converting the sff. to fastq using mothur, then running split_libraries_fastq.py in Qiime, but because this is 454 data, not Illumina, I don't have separate barcode read fastq files, so the script won't work.

Can anyone help me with this?

[coralnerd]

Never mind - Qiime has a script that solves my problem nicely.

make_fastq.py will combine the .fna file output of split_libraries.py with a .qual file to produce separate .fastq files for each sample!

[mcnelson.phd]

You should also look into using make_per_library_sff.py. That will take your original sff file and, as the name implies, make sff files for each sample. That way you're uploading your data in the original sff format that 454 data is provided in and not processing it and then uploading files that are a few steps removed from the original data.