Split already joined and demultiplexed pair end sequences into separate FASTA files

Question

[haig.they]

I have MiSeq Illumina sequences that have been already trimmed (for barcodes and primers) and denoised with dada2 in qiime. All my sequences come from two samples that are mixed up inside a single .qza file. How can I split my sequences in two samples according to the sample-metadata into two FASTA files?

[lieven.sterck]

First, what is a .qza file? I personally have never heard of it.

so you sequenced a pool of two samples in a single lane, correct? if so you can only split out the samples by using the barcode for each sample in the pool. Did the sequencing provider did not already demultiplexed them for you?

[haig.they]

Thanks for your response!

.qza is a qiime artifact (they are fastq imported files). My samples are already demultiplexed and the barcodes have been removed (done by the sequencing provider). After the dada2 step, all sequences from the two samples were merged together in a single file. I just want to split them back together in two according to sample name.

[GenoMax]

Are the sample names present in fasta headers?