It's rather annoying to have to sort the DNA and AA sequences for back translation, given that alignment on the AA seqs is likely to lead to reordering. I get that for larger sequence sets, this restriction let's us more efficiently iterate over the data, but I'm doubtful this frequently outweighs the inconvenience of having to reorder sequences yourself. It would be pretty easy to add an optional flag that sorts the sequences for you by seqid (or by dna order or aa order, perhaps also controllable via option).
It's rather annoying to have to sort the DNA and AA sequences for back translation, given that alignment on the AA seqs is likely to lead to reordering. I get that for larger sequence sets, this restriction let's us more efficiently iterate over the data, but I'm doubtful this frequently outweighs the inconvenience of having to reorder sequences yourself. It would be pretty easy to add an optional flag that sorts the sequences for you by seqid (or by dna order or aa order, perhaps also controllable via option).