Quick guide to Vienna RNA fold

  • 3
  • Article
  • Updated 7 years ago
1)
Open Vienna. You could end up using it a lot, so you might want it in the bookmarks of your search engine.

2)
You get the RNA sequence for the a design, by clicking on this copy page symbol in the tool section in Eterna.



3)
You paste the RNA sequence in this field in Vienna.



Then click this bottom



4)
What to use the numbers for.

The MFE structure number should be high, as it means your design is more likely to fold correct.

Ding said that ensemble diversity should be below 0,5. She also said that entropy should be under 0,3. These are good tested rules.

However lately we have had designs with triloops in them. In those design entropy and ensemble diversity will be higher. That is also the case, to a lesser degree, in designs with asymmetry. But remember - low is good.

5
How to read graphics.

Notice this section under Grapichal output. Move the dot from base-pair probabilities to positional entropy. This will give you a better view on where there might be weeknesses in your design. A red colored design means that your design is good or that it is a christmas tree - which is bad. An all red design is rare, so just try to turn cold colors in to warmer colors. (Without using too much GC-pairs when designing) Vienna is your chance to get an idea about where you need to change things.



Here is a picture of the entropy of a design. This one will not do good. It has 5 spikes with high entropy. You need as few spikes as possible and you need them low. The numbers from 0-100 represents nucleotides in the design. You can use this to see where in your design entropy is bad.



If you want to know more about Vienna RNA fold and especially about how and why it works, you should check Alan.robot's introduction to Vienna. This quick guide is just an easier version of his page about how to use vienna.
Photo of Eli Fisker

Eli Fisker

  • 2141 Posts
  • 457 Reply Likes

Posted 7 years ago

  • 3
Photo of Jeehyung Lee

Jeehyung Lee, Alum

  • 708 Posts
  • 94 Reply Likes
Eli this is an awesome guide!

This should definitely go into user strategy guides list if that's ok with you
Photo of Eli Fisker

Eli Fisker

  • 2141 Posts
  • 457 Reply Likes
Hi Jee

Thanks!

That would be great with me
Photo of Eli Fisker

Eli Fisker

  • 2141 Posts
  • 457 Reply Likes
Addition to my guide:

I said Ding's rule for entropy under 0,3 was great. But if you design like Mat, with all A's in tetraloops, then you should accept higher entropy numbers from Vienna. Up till 0,6 is propably okay.

Another really important thing. Clollin, Starryjess, Mat and I have been discussing the use of different RNA tools lately.

Mat sent me this link today. It is an exceptional post by Dimension9 and Mat from long time ago, that deserves new attention. It holds valuble information on how to use Vienna. Vienna is the best tool we have so far and I love it. But there is some things you should know, to get the best out of it.

Things learned from the use of Vienna RNAfold

Mat said: I think the newer players are using the online tool as is and not as a guide only.

I think that cuts right to the core of the problem with our use of tools. Even I as experienced player have jumped into this trap.

If you don't believe him, check his score average - and he don't even use Vienna. :)

I for long have used the rules for designs that we have found, to discriminate between good and bad designs. Then I consult RNAfold. As an extra.

Notice what Dimension9 says in the post: I have a feeling that if we are to have a truly predictive tool for synthesis success, we will have to develop it ourselves.

I think these words are very true.

And as Clollin said: I think any one tool or even metric within a tool is not enough
- u have 2 verify all aspects of many tools and see a trend occurring.

So until we have our own homemade tool, use the tools available, but be sceptical.
Photo of rstager

rstager

  • 4 Posts
  • 0 Reply Likes
To me, it seems that our objective should be to build better algorithms and better tools. What I am observing is a hybridization of man and software that is allowing us (you) to see new characteristics that are important to how the molecule folds that were not readily visible to the scientist or the programmer. As such, enhancing the tools to improve visualization and usability should be paramount.

It would also seem that visualizing and comparing the lab results is more important. Ultimately we want to build sequences that work with nature, not ones that work with our existing software models.

I'm a newbie to EteRNA, so please forgive me if my comments are naive or misguided.
Photo of Eli Fisker

Eli Fisker

  • 2141 Posts
  • 457 Reply Likes
Hi Rstager!

I like that you make it so clear. When we using the existing RNA tools, we become man/machine hybrids. :) And not using our precious brains unfiltered.

I won't personally throw all tools out, as I recon them useful to see if ones aim is totally off the target. They are also helpful in the learning process about what might work. They can visually show that a cub scout design won't work, although they don't tell on christmas tree designs.

Sometimes the RNA tools even works as inspiration for improving the game.

My play with the Genebee tool (which is no good what Eterna is concerned) made me realize, that the winning designs had an overall even energy distribution. Something we could not see back then, as we did not had the energy visible ingame. I wrote the post Energy structure and symmetric colors, after which Mat requested that we should have an energy tool, showing us the energy between the stacks in the designs and puzzles we were working on. Our wish got fulfilled.

But as Jee just said about the free energy numbers in the game: they are useful - but also it's important to know that those are numbers that aren't 100% correct

So your point is taken. Keep up the good work.
Photo of Eli Fisker

Eli Fisker

  • 2141 Posts
  • 457 Reply Likes
I just want to make a link to Quasispecies beautiful graphs on determine the relationship between score of designs and ensemble diversity.

How useful is structure predition software in lab

So you might want to accept a bit higher ensemble diversity than the 0,5 limit Ding and I set before.

But some of the tendencies I mentioned hold. In asymmetric designs the highscoring designs will have higher higher ensemble diversity than the symmetric ones.