Hello all. Nice to be here. As a newbie I have 2 questions: (1) When solving puzzles there are many solutions - some with lower energies (desired). Does this matter or is the only main objective to get the fold right? (2) Some puzzles have restrictions, e.g. no more than 5 C-G bonds, etc. What is the purpose of these restrictions? Is it a reflection on some natural requirement or is it arbitrary? Thanks!
1 Like
OK, so I’m replying to myself - LOL. OK, I got question 1 down after reading other posts and the fine strategy guides from Brourd, Fisker, et al - THANKS! For general recreational puzzle solving you can just get the RNA to fold correctly. But in real life you want to make the fold as energy stable as possible - so I suppose when it gets to lab-based puzzles for synthesis (competition) then we want the best most stable designs - low energy.
So, only question #2 I’m not so sure about.
As my knowledge evolves, again through the help of others, I’m beginning to answer my own questions. I’m documenting it here in the hopes it might help someone else. The goal of lab puzzles is NOT to simply get the highest free energy as I mentioned above! Too little free energy and the bonds will be too weak to be stable and too much free energy means the RNA might NOT fold, IN REAL LIFE as expected. So now I’m beginning to see that there is some art needed here in all of this.
Yeah, but what is the golden median between kcal and stability? You guys in the lab - we want to get inside your head - spill the beans… There must be a way to do it perfectly…
Yo 0bodyCount, Fisker’s Lab Guide for New Players has some interesting tips regarding energy balance and other tips:
Nice one, thanks
Actually YoHobie - Go to the archives and check out the winning designs. That says it all. Its about perfect chaos in perfect randomness in a symphony of balance. The best designs look like when you cross a (half Norwegian/half Jamaican) with a (half Japanese/half African) = the ultimate bastard mutt…
As I see it now (after having read a shitload), the goal is to not give the RNA any chance to find any other form, and to this end it seems the best way is to strive for endless uniqueness. To not allow any repetition. Attached is the last one I made. But when it comes to loops, I’m still a bit loopy for lupins. I try to randomize, but I get the drift from the textbooks, that the lowest kcal is not always the best for stability. Perhaps a balance between the sizes, position and kcal of the loops also needs to be achieved… 
So now I first drop the hard points, then randomize AU between them, then boost and decorate the loops and make sure they dont loop themselves out of the loop. Then randomize GC on top of the AU, then sprinkle with GU. But when I use GU last, it always completes the puzzle when I reach the target amount. Perhaps I should apply GU earlier in the process in order to overshoot and achieve greater chaos…Bear in mind, that reaching the GC count is not a must either, but also a matter of balance… 
Looking at these winning designs is very instructive. Thanks.
I like what you said above. It’s funny when I first started the simplistic idea is “G-C bond strong…ooooh good strong bond, use plenty of G-C bonds all over”! How hard is it to do these puzzles? LOL. Then you advance to maybe using G-C bonds only at the end point of stacks and then use other base pairs to fill the middle plus throw in some random changes to minimize repetition - and make it look pretty. But these are all blind procedures - there is no art in that! Then I think when you submit your design to get a dot plot or melt plot you realize nature finds WAYS AROUND WHAT YOU WANT and that only by coaxing her with energy wells strewn about will it fold correctly. Folding after all is dynamic and subject to perhaps intermediate states and environmental influences, etc.
1 Like
I went through the same stages. I remember in some of the big designs I could coax an RNA to fold correctly by using more GC at a specific point and if I replaced them with AU it would unfold again. Thats like a brute force method. GC has its uses especially in loops but the two other bonds are just as important in order to achieve uniqueness…
A new thing thats boggeling my mind now, is where is the best place to start the build. Logic suggest it would be dead in the middle of the string?? … to put a pinch on the whole thing… Any thoughts…
0bodycount, I’m not yet ready to say anything meaningful at this point. I’ve been digesting the various strategy documents and forum entries and accumulating a list of rules to follow. But more than just having rules I am trying to determine when it is appropriate to apply them. Do some rules apply best to static versus switch puzzles? In general is it better to have less G-C bonds and more U-A bonds? What do the winning designs show - individually as well as by “categories”? Should you boost loops first and then stabilize the stacks or the other way around - all of one or all of the other - OR BY SECTIONS? What sections are best to start? One player suggested a go slow approach so as not to spend time undoing things and that the best process is to size up a puzzle and know where to start! And then the big question for me is what are the major differences when solving challenge/player puzzles versus lab puzzles?
I guess thats why they had momentum to create this webpage and this game. If it was all a walk in the park, we would all be sitting in the park by now sipping icecold RNA. I guess all these unknowns are what makes it exciting and gives us room to experiment…
Glad to see some constructive discussion going on around here. 
The restrictions are to help strive for balance. For example, you don’t want to have too many GCs, or else there will be a lot of attractions going on you don’t want (they’re strong enough that they may even cross the RNA). Though, at the same time, keep in mind you want enough to keep the design stable.
EDIT: Also, there are some restrictions saying that you can’t have a certain number of bases in a row. This is often because it can screw up synthesis a lot… See here for some recent issues: http://eternagame.org/web/blog/4147105/