Allowed these patterns in small multiloops (3 arms, one neck, and no nucleotides between the (4) gc-pairs)
a)
Rightturning GC-pairs all around allowed (red nucleotide to the right)
b)
Axis I: The GC-pair in the neck turn opposite, so does the one in top, the last two turn right (red nucleotide to the right), the rest turn the usual way.
c)
Axis II: Neck GC and top GC-pair (with red nucleotide to the right), left GC pair and bottom GC-pair opposite.
For any other combinaiton give -1 for each GC-pair in the multiloop that is differing from these three patterns.
See the reason for this in my getsat post loop pattern for small multiloops
This is sort of meant as an adittion to more general strategies for GC-pairs in multiloops. Could be seperated into 3 different additions.
Dear Eli,
Your strategy has been added to our implementation queue with task id 27. You can check the schedule of the implementation here.
ETA of the implementation is 6/22/2011
Thanks for sharing your idea!
EteRNA team
Hello Eli,
I’m Jerry Fu, I’ll be helping implement the script for your strategy. I had a few questions to clarify the logic.
-
How would you like to treat multi-loops with stacks that are non-GC?
-
In the images you provided, I deduced that the neck was right and the arms were the other directions. Is this correct? I also deduced that “right” was if you rotated it so that the stack we’re looking at is at the top and the multiloop is below it.
-
If an RNA has multiple multi-loops, and say one scores 98 (2 pairs incorrect) and the other scores 96 (all incorrect direction), should the final score be an average (97), lowest (96), or subtract total? (ie. Final score would be 100-2-4=94)
Thanks for the help!
Also, sorry for the delay in getting this out! There were a few problems with the script submitted on 6/22 (my fault), so it’s been pushed back a little bit. Hopefully this won’t happen again!
Hi Jerry!
-Yes the neck was to the right, like you correctly guessed. (my fault)
-Yes, right and left in the picture is from when looking on it from the top. And twisting your head around 
-With right (correct) direction for a GC-pair in a multiloop, I mean Red nucleotide to the Right. (my little memory rule) And green to the left.
Basepairs in the multiloop ring that are not GC-pairs, I sort of don’t like. All our experience tells us that there have to be GC-pairs in junctions. So what would be best? To rate them the same as or just -1 worse than if the GC-pair turns opposite compared to where I want it? Think I go for the last option. A missing GC-pair is not a wrong turning one.
So if an opposite GC-pair gets -1, give a missing GC-pair -2.
I have a question:
What do we do with the designs that have no small multiloops? Those two picture patterns (beside the usual one) I ask for in this strategy will only be occouring in the small mulitloops with 4 GP-pairs and no nucleotides between the arms. And will only appear in puzzles like The Cross and The Bulged Cross. This algorithm is sort of meant like an addidion to a more overall algorithm for multiloops. Could it be made to rate only those puzzles where it is relevant?
You asked about score for puzzles with more than one multiloop:
If an RNA has multiple multi-loops, and say one scores 98 (2 pairs incorrect) and the other scores 96 (all incorrect direction), should the final score be an average (97), lowest (96), or subtract total? (ie. Final score would be 100-2-4=94)
Then final should be 94. The more wrong turning GC-pairs, the less negative energy the multiloops have inside to keep the structure together and the worse the design will be.
Good luck!
Hey Eli,
Thanks for your quick reply. I’ve already implemented the specifications you asked for and it should be in review before being published.
About the question you asked, yes, it is possible for our functions to decide that certain designs are “unscorable”, in which case it will receive a “N/A” score and not effect any of the final rating calculations.
Thanks for your help!
Dear Eli Fisker
We are glad to report that your strategy has been implemented and tested.
While implementing your strategy, we have made small changes to the parameters you specified to optimize the performance.
Note that we’ll always run a optimization over the parameters you specify, so you won’t have to worry about fine tuning all the numbers you use.
Just the idea and rough numbers are enough to run your algorithm!
Length : Your strategy was implmented with 65 line of code.
Ordering : We ran your strategy on all synthesized designs and ordered them based on predicted scores. The correlation of your strategy’s ordering with the ordering based on the actual scores was 0.0132107450489. (1.0 is the best score, -1.0 is the worst score. A completely random prediction would have 0 correlation)
Please note that the numbers specified above will change in future as we’ll rerun your algorithm whenever new synthesis data is available.
More detailed result has been posted on the strategy market page. Thank you for sharing your idea, and we look forward to other brilliant strategies from you!
