Bots should follow the same restrictions as we are given in player puzzles. That’s all.
Freywa
they can’t solve most of them even without the restrictions - perhaps adding the restrictions will improve their performance by elimination of ridiculous strategy attempts in their iterations ?
I don’t think this is possible. RNAinverse (Vienna’s inverse folding program) and the web version of InfoRNA don’t seem to allow it. Not sure about RNASSD.
I don’t see minimum or maximum values for certain kinds of pairs listed among the constraints you can specify. Though I do think you could do fixed bases.
And as I said before, I take the same view, but skewed. If the bots don’t have to follow restrictions, neither should we, in the labs at least. The stabilized plot should be a suggestion, not a requirement. I haven’t noticed if bots have the multiple-bases-in-a-row restriction, but if they don’t this should also be a suggestion, not a requirement. In the puzzles, the bots are so hopeless maybe we should just consider the restrictions a handicap for we humans who are better players! Maybe for the puzzles where bots ignore restrictions we shouldn’t say they solved the puzzle, only that they solved the shape (Big deal, I have a number of white solutions where I can’t get the last couple GUs in or the last couple GCs out.)