A very important area of nanotechnology--self-assembly--is addressed in this paper. It considers several algorithmic approaches (exhaustive branch and bound, heuristics with some relaxed objectives, and answer set programming) to find a minimal (or near minimal) tile set for any pattern. Each algorithm has its own merits and demerits, but there is no discussion on the implementation aspect of the problem.

There must be some design study on the wet lab implementation of the different tile sets. It would be nice if there were some indication of which of the three algorithms would be more cost effective from a design point of view.

This is a very technically sound paper. The overall treatment is very good, but it would be more attractive, from a practical point of view, if it discussed the wet lab implementation aspects.