I would like to suggest a more friendly way of allowing
mummer to interact with scripts.
First of all, I am frustrated because mummer needs to
read imput from files and that you cannot pipe
sequences directly to mummer. Could that be fixed
somehow, so that if you pipe fasta files to mummer, the
suffix tree is build from the first sequence, and that
subsequent sequences are used for matching? (I am dying
to get this functionality).
Alternatively, some sort of API may be worth considering.