In MiSeq SOP, align.seqs is run before classify.seqs. I am wondering whether I can run classify.seqs before align.seqs. It could save some CPU time in the classify.seqs step since my input reads are unaligned. I guess the order does not matter for final results?
Alignment doesn’t affect the classification. Actually waiting until after you align, screen, unique, pre.cluster, etc. and remove chimeras will result in a smaller dataset that will actually be easier to classify than the original dataset.