[dekhn]

I've explored alternatives to FASTA and FASTQ but in most cases I found that simply not storing sequence data is the best option of all, but if I have to do it, columnar formats with compression are usually the best alternative when considering all of (my) the constraints.