What do you think of the competition puzzles?

The development team wants to know what you think of the competition puzzles.
Are they fun?
Can you think of any ways to improve them?
Any other comments?

Could you explain/show how the information/data is being used and how the data collected will benefit the game ?

I like the competition puzzles a lot.

Edit: the thing I requested appears to already be in place; we can comment on a specific competition puzzle. :smiley:

This is just an informal survey for the development team to gauge how the players feel about the competition puzzles. If you or someone has a good idea about the topic, it might be implemented into the game.

  1. Add new types of competitions.
  • G-C competitions (the less GCs the better)
  • combined GU/GC competitions
  • lowest/highest energy
    all of these with different other conditions
  1. Add a table “Solved Competition Puzzles” to the profile page, containing the columns:
  • puzzle (linked)
  • rank
  • of players with the same rank

  • of players who solved the puzzle

  1. Add a dynamically calculated “Competition Points” (CP) leader-board where you can earn CPs by ranking high in competitions
  • example of a points scheme:
    #1 = 30 CPs, #2 = 20 CPs, #3 = 15 CPs, #4 = 10 CPs,
    #5 = 8 CPs, #6 = 6 CPs, #7 = 5 CPs, #8 = 4 CPs,
    #9 = 3 CPs, #10 = 2 CPs, #11 - #n = 1 CP

Yes! Yes yes yes! To both 1 and 2!

Except maybe the points leaderboard. Everyone who does the competitions knows who else is good at them. :slight_smile:

Just a thought, and I don’t know the science behind this well enough to know if it’s feasible, but…

One thing I constantly find myself wondering about is how my solutions compare to the “real world” versions. I’d be interested in seeing a set of constraints to match real-world situations.

So for instance, if a certain RNA is known to be 30% G, 30% A, 20% C, 20% U try to match those numbers rather than just minimize GC pairs or maximize GU pairs.

It could also be done with proportions of bond types instead of individual bases.

I also think this might be helpful as practice for the lab. I love the challenge puzzles, but am not sure that the intensive practice in trying (for example) to squeeze all the GU’s I possibly can into a design is as useful training-wise as a focus on more balanced designing might be.

lol 5c, yes indeed, madde :stuck_out_tongue:

I don’t know what the devs have in mind for future lab challenges, but I think in preparation for the post “one-bulge-cross” era, puzzles should involve different design constraints than are currently used. Overall % CG would be an easy one, and “locking” sequences around functional regions (i.e. known to bind ligands or catalyze reactions or form tertiary interactions) would be the logical next step.

The nice thing about GU competitions is that it forces people to get to know the loop and stacking energies really well, but it probably isn’t representative of the types of design constraints you’ll be exploring in the future.

Instead of having separate puzzles in a series with different levels of constraints and corresponding rewards, just have one puzzle and score it based on which of those sets of constraints it meets.

I’m curious about why we compete for GU count, is that in some way useful, does it match something seen in nature?

some answers about GUs at the bottom of this thread:

alan, I see, the short answer being that the GUs do occur in nature and the scoring algorithm doesn’t reward for them.

That would seem to be because the scoring (energy) algorithm is one-dimensional, tracking benefits without costs. If there were a cost dimension offsetting energy benefits, the low cost of GUs might be realized in a single metric.

I realize this can hardly be a new idea, and the development of many dimensionalities of scoring are all research topics - and you did answer my question! I’m just going to have to go get some actual texts on this stuff, if I want to know much better!