Please vote for the lab shape you would like to test next (after “Bends Ends Sampler”)
We would like to ask you to pick puzzles that can expose the weakness of the current energy model and algorithms. In other words, we want you to pick a shape that bots such as ViennaRNA and NUPACK will fail!
This poll will close on July 4th, Monday - We are looking forward to seeing another interesting lab challenges from players!
I think the sharp angles on “Almost a parallelogram” by Brourd would be good to test. “Branching Out” by Brourd might also be interesting and neither ViennaRNA nor RNASSD was able to solve it.
Hard Y and a bit more is too hard, I think. The lab puzzle needs a bit of leeway. I would prefer a design where RNASSD failed but 20 or so humans succeeded.
I second all these nominations. What I particularly like about Broken Propeller and Tilted Picture of Running Man is that they have extra-long unbonded ends which I think will help us get good data about how open loops work.
If we end up with Human Astrovirus, I think it would be really interesting to reserve one slot for the “real-life” sequence, for comparison’s sake.
he present puzzles are too complex in design, with either not enough time given to diagram one properly, and the branches carry on forever, or else the winning aspects of a puzzle are not clearly defined in their conception. I offer this criticism knowing it is very difficult to create one of these puzzles, being the web site is visited by experts as well as amateurs as myself. It is a real privilege for me to be able to participate in these games.