I decided to discontnue support for the UTAX algorithm because on balance, I think SINTAX is better.
SINTAX has two main advantages: it does not require training, and it works better than UTAX on non-16S genes.
The main advantage of UTAX was that it was (still is) the only taxonomy
algorithm I'm aware of which does not tend to over-predict on novel 16S
taxa. But this is achieved, inevitably, at the expense of somewhat lower
sensitivity to known taxa, so even on 16S the advantages of UTAX over SINTAX
are not very strong.