User:ElNando888/Paper Lab: Riboswitches

----

Intrinsic stability hypothesis: R91-based candidates

4/4 base pairs 5'

GGAAAAUAAAACAAAAUAAAUAAAAUACUCCCGGAAGGCACAUAAGUGCGAAGUGUGGUCCGGUAGGACCAUCAUCGUAUGAAACGGGAGGCCUAGACUUUCGAGUCUAGAAAAGAAACAACAACAACAAC

-FMN	+FMN(200)	Mimic

5/5 base pairs 5'

GGAAAAUAAAGAAUAAAAGAAAAGGCGCACAAAACACGGGAAACCGUGGAAUACUGGGUAUGACUACCCAGAAUCUACGGGAAAGUGCGCAAGGUGGGCUUCGGCCCACCAAAAGAAACAACAACAACAAC

-FMN	+FMN(200)	Mimic

6/6 base pairs 5'

GGAAAGAAGAAAACAAAAUAAACUCAGAAAAGACAAAGGAAUUUGUCGUUGCGUGCAGCGGAGGCUGCAGGCACGAUAGAAGAAUUGAGUAACCAGGUGUUCGCGCCUGGAAAAGAAACAACAACAACAAC

-FMN	+FMN(200)	Mimic

7/7 base pairs 5'

GGAAAAUAAAAGAUAAAAUGGCUCAAAAGACACGUCGAGAUGUGUCCAUUAAUCAGAGGAGACUCUGAUCCUGGAUAUAUAAAAGAGCCAGAGCUCGGGUUCGCCCGAGCAAAAGAAACAACAACAACAAC

-FMN	+FMN(200)	Mimic

4/4 base pairs 3'

GGAAAAGAAUAAAAUAAAAGAACAAAGCGCGUGAAAAGUGGCUAAGGUAGUCCAAAAGGACUACAAUGCCGCGUAAGCGGAAAACACGUGAAGGCCUCCUUCGGGAGGCCAAAAGAAACAACAACAACAAC

-FMN	+FMN(200)	Mimic

5/5 base pairs 3'

GGAAAUAUAAAACAAAACAAAACACACCGGAAAAGGAAUGAAACGUCCGCUGAAUAGUGGAUGCUCGUUCCGGAAGGAACAAAACUGGUGAACUGCCCUUUCGAGGGCAGAAAAGAAACAACAACAACAAC

-FMN	+FMN(200)	Mimic

6/6 base pairs 3'

GGAAAAGAUAAAAUAAAAGAGCUGCCGAAAAGUAAAUCAAGUGUCAGACUACGGUCUGAUCGUGGUUUACGGAAGUAAGCAAAACGGCAGAAGUUGCCAUUCGUGGCGACAAAAGAAACAACAACAACAAC

-FMN	+FMN(200)	Mimic

7/7 base pairs 3'

GGGAAAAGAGAAAAGAAGGUACUCGAAAGAGUUAUGUAAGUUCGGAGACACCUCCGAAAAGCGUGAUUUGGAAGGAUCACAAAGGAGUACAACCAGGGCUUCGGCCCUGGAAAAGAAACAACAACAACAAC

-FMN	+FMN(200)	Mimic

Shift 4/4 base pairs 5'

GGAAAAGAAAAGAAAAUAAAAGAAAAGAAGACCCUAAAGCUACGACAGUAGAAAUUGGUGCAAUAGCAUCAAAACUAUUGGAAAAGGGUCAAGCUCUCAUUCGUGAGGGCAAAAGAAACAACAACAACAAC

-FMN	+FMN(200)	Mimic

Shift 5/4 base pairs 5'

GGAAAAGAAAAGAAAGAGAAAAUAAAGGGCCUCGAAGCUGACAGACGUCAGGUAGUUGACGGGAACGUCGAGGACUGAUGGAAAGAGGCCAAGCUGUCAUUCGUGACAGCAAAAGAAACAACAACAACAAC

-FMN	+FMN(200)	Mimic

Shift 6/4 base pairs 5'

GGAAAAUAAAACAAAUAAAAUAAAAUGCGGCAAAAUUCUGUACAAGCAGAGAUCCUGGCGCUACGGCGUCAGCACUUUGCAAAGGCUGCAGAGUUGGCAUUCGUGCCAACAAAAGAAACAACAACAACAAC

-FMN	+FMN(200)	Mimic

Shift 5/5 base pairs 3'

GGAAAAUAAAGAAAAUAACAAAGAAAGCUUACAAAGAUGCGCUAGACUGGCCGGCAGUGGCCAGAGGGUGCGUAAGUACCAAAGGUAAGCAAGGUGGGCUUCGGCCCACCAAAAGAAACAACAACAACAAC

-FMN	+FMN(200)	Mimic

Shift 6/5 base pairs 3'

GGAAAAUAAAUAAAAGAAUAAAGACUGGGCAAAAGUUUACUCACGUGACGAACAAUCGUCAUGAGGUGGACGGCUCUACUAAAAGCCCAGAAGGAGUGCUUCGGCACUCCAAAAGAAACAACAACAACAAC

-FMN	+FMN(200)	Mimic

Shift 7/5 base pairs 3'

GGAAAAUAAAACAACAAAAUAACCCACUAAAAGAGAUCAAAGUUCCAGCGCAAGCUGGAAUAGGAUCUCGCUCGAGGUCUAAAAGGUGGGAACCCGGUCUUCGGACUGGGAAAAGAAACAACAACAACAAC

-FMN	+FMN(200)	Mimic

How did these score?

Salish99 (talk) 09:49, 8 December 2014 (UTC)

These have been neither synthesized, nor scored yet. I'm planning on creating a pilot lab where I would re-use the same targets as were tested in round R91, but this time, I'd like to test a personal hypothesis which essentially states that the FMN aptamer may have an intrinsic stability that is (possibly vastly) underestimated by the Turner models.

ElNando888 (talk) 11:21, 8 December 2014 (UTC)

R86 Analysis

R70 references

The first questions:

how well do mutations in the binding site mimick the effect of the presence of FMN?
how do we predict the 'fitness' of a given mimic

Example: Hair Trigger

This is the baseline, the original design and result from R70

And this is one of the -2.51 kcal (vrna211) mimics tested in R86.

In theory, the best mimics are those that produce the exact same SHAPE signature over a maximum of bases, ignoring the binding site (and probably the barcode too)

TODO:

for each tested mimic compute the RMSD to the R70 baseline, and sort the list
tune the fitness algorithm so as to reproduce the same ranking order as best as possible

New targets

(...)

----

Thermodynamic Mimics: A Method for Simulating Functional Riboswitches

Mimic-related scripts

The scripts are on a dev server, because some libraries are not available on the prod server yet.

http://nando.eternadev.org/web/script/

Disclaimer

These scripts are based on my uneducated understanding of nucleic acids, thermodynamics and multi-state RNAs. Moreover, they probably look embarrassing to experienced coders' eyes, For my defense, I can only say that javascript doesn't belong to my professional practice, and I was in a hurry...

I'm providing these scripts in the hope that they will prove useful. Everyone is welcome to discuss, copy and/or modify them.

Constraints lines

The provided scripts use an unusual (and perhaps confusing) input in the form of a ViennaRNA-formatted constraint line. These chains of characters have the same length as the considered sequence, and each character represents the desired pairing state for the corresponding base:

for bases that may be either paired or unpaired, use a dot (.)
for bases that must be paired (no matter with which other one), use a pipe (|)
for bases that must stay unpaired, use a small 'x' (not a capital one)

The reason why I introduced their usage is economical. Given the sequence and the constraint, the scripts can:

get the offsets for the binding site, both 5' and 3'
infer the alternate sequence (by replacing the mimic by the FMN aptamer and vice-versa)
infer the secondary structure of the OFF state (simply by folding the FMN-aptamer version)
infer the secondary structure of the ON state (simply by doing a constrained fold)

In conclusion, it allows to give only 2 inputs to the scripts (sequence and constraint) where no less than 5 would be needed (sequence, OFF structure, ON structure, binding site offset 5' and 3')

An additional and less obvious advantage is that since secondary structures are implicit, it allows for the scripts to accept "normal" designs, as well as what some deluded designers would possibly attempt to use under circumstances: designs that include non-canonical pairs in either or both states.

Free energy effect of a mimic

	OFF state	ON state
FMN aptamer	A	B
Mimic	C	D

In the above table, columns are related to secondary structures, and rows are about the sequences.

A and B are the switching states we are used to deal with. The difference FE(B) - FE(A), where FE(X) means "free eneergy of X", would be this delta that we need to keep under the usual -4.9 bonus (and which we would like to reduce to -2.51).
C's structure is exactly the same as A's in the ideal case, but due to the mutations caused by the mimic, some pairs need to be added to and/or removed from the structure.
D's structure is normally identical to B's, with the expected exception of the pairs that are created inside the binding site.

In this constellation, the global effect of a mimic is simply computed like this:

ΔG(mimic) = FE(A) - FE(B) + FE(D) - FE(C)

By the way, what we're doing in this lab, is to:

create a switch
select or design its mimic
submit (A) in the "OFF state" lab
and (D) goes into the "Mimic" one.

Mimic evaluation

Example with EteRNA's model (ViennaRNA 1.8.5), in bold, the unpaired bases of the aptamer binding site:

GGACAAAAUAUGAGGAUAUAUCGAAAGAUAGAAGGCAUAAAAUAGAAAGCUUUCAAAAGAAGACUGACUUCGGUCAGUCAAAAGAAACAACAACAACAAC

(note: delta free energy is 2.2 kcal)

We will also take the following mimic sequence as example:

ACACCA--GGAAG

In target mode:	In natural mode:

Notice that considering only the free energies, the effect of the mimic seems to be an improvement of -2.3 kcal.

For inputs, we need first the sequence with the mimic (in bold)

GGACAAAAUAUGACACCAUAUCGAAAGAUAGGAAGCAUAAAAUAGAAAGCUUUCAAAAGAAGACUGACUUCGGUCAGUCAAAAGAAACAACAACAACAAC

We require the description of the position and size of the binding site in the form of a coded constraint:

...........|xxxxxx|..........|xxxxx|................................................................ (need to explain how to get this)

Finally, we need to indicate what binding bonus we're trying to "copy", and the following value matches a FMN concentration of 200 uM, as was used in R70.

-2.51 (see my blog post about aptamers and bonuses for more details)

(someone please tell me what's not clear here?)

Mimic search engine

Example stolen from a puzzle solution:

GGAAAAAAUGCAAGGAUAUUACGAAAGUAAGAAGGUGCGAAAAAAAAAACCUUCAAAAAAAGUAGUACUUCGGUACUACAAAAGAAACAACAACAACAAC

The free energy delta between states is 1.9, which is suitable for a -2.51 mimic.

The first 3 inputs are very similar to the ones for the evaluation script, the only difference being that the sequence should be the one containing the FMN aptamer.

The remaining input is related to the search algorithm, and defines a number of iterations. To explore a 4 million units search space, 10000 sounds reasonable, and on my machine, it usually results in waiting for 2 to 5 minutes depending on the length of the sequence, which I find acceptable. Tune according to your preferences.

Better set the timeout to a high value...

Depending on input parameters and CPU speed and availability, this may take a couple minutes, up to a couple hours...

If everything goes well, the script will finally spit out a long list of scored mimics, which will be the best ones it could find within the allocated number of simulations. Ignore the first mimic over the line, it's just the last "branch" that the engine was exploring at the time it stopped.

Since this is a Monte-Carlo search algorithm, this list will never be exactly the same. If you're getting too few mimics to choose from, or not good enough, try increasing the number of simulations and re-run.

I ended up selecting (a bit randomly in this case) the mimic marked in green.

Target mode	Natural mode

----

FMN Switch Mimic Pilot 1

Candidates

Meechl 1

Predictor	OFF state	ON state	ΔG	Notes
ViennaRNA 1.8.5 (EteRNA)	-30.63	-27.70	+2.93	The ON state is predicted to have 2 additional pairs by ViennaRNA 2.1.x
ViennaRNA 2.1.x	-24.00	-23.10	+0.90

JR 1

Predictor

OFF state

ON state

ΔG

Notes

ViennaRNA 1.8.5 (EteRNA)

-35.93

-33.30

+2.63

ViennaRNA 2.1.x predicts that the ON state will be the MFE.

It would have been preferable to break the poly(A)s. If there are enough candidates before the deadline, this may be the reason why this design could eventually not make the cut.

ViennaRNA 2.1.x

-28.00

-28.20

-0.20

RedSpah - Babel

Predictor	OFF state	ON state	ΔG	Notes
ViennaRNA 1.8.5 (EteRNA)	-30.33	-26.70	+3.63	The OFF state is predicted to be missing the entire neck stack by ViennaRNA 2.1.x
ViennaRNA 2.1.x	-25.30	-22.90	+2.40

ElNando888 - Meemeec

Predictor	OFF state	ON state	ΔG	Notes
ViennaRNA 1.8.5 (EteRNA)	-37.63	-34.60	+3.03	Both states feature non-canonical pairs.
ViennaRNA 2.1.x	-32.20	-30.70	+1.50	Both states feature non-canonical pairs.

----

Designer's comments:

It has been shown that non-canonical pairs work quite well in single-state designs. It remains to be seen how these pairs will behave in the context of a multi-state one. It is hoped that they will provide flexibility, without sacrificing too much in stability. -- ElNando888 (talk) 04:24, 27 March 2014 (UTC)

JR 2

Predictor	OFF state	ON state	ΔG	Notes
ViennaRNA 1.8.5 (EteRNA)	-25.63	-23.70	+1.93	ViennaRNA 2.1.x predicts that the ON state will be the MFE. Poly(A)s still unbroken, same comment as for JR 1.
ViennaRNA 2.1.x	-19.10	-20.00	-0.90

JR 3 (removed)

Predictor	OFF state	ON state	ΔG	Notes
ViennaRNA 1.8.5 (EteRNA)	-28.53	-27.50	+1.03	ViennaRNA 2.1.x predicts that the ON state will be the MFE. Poly(A)s still unbroken, same comment as for JR 1.
ViennaRNA 2.1.x	-22.30	-23.90	-1.60

Vinnie 1

Predictor	OFF state	ON state	ΔG	Notes
ViennaRNA 1.8.5 (EteRNA)	-40.73	-36.90	+3.83	.
ViennaRNA 2.1.x	-36.70	-35.10	+1.60
RNAstructure	-36.40	-34.90	+1.50

----

Out of curiosity, I tested this sequence with RNAstructure:

http://rna.urmc.rochester.edu/RNAstructureWeb/Servers/Fold/ResultsPages/832160601136798877/Results.html
(I'm not sure how long this link will last, hopefully long enough)

The results are in very good agreement with those of ViennaRNA 2, and RNAstructure seems to think that those 2 structures are the only major ones for this sequence.

This dynamic aspect of the switch function will not affect the testing with mimics. Though, and for the sake of mentioning it, the barriers webserver from ViennaRNA seems to show that the sequence would face a rather serious energy barrier problem, if it were to be tested with an actual addition of FMN.

http://rna.tbi.univie.ac.at/barriers/gJUKF3JE52/

Vinnie 2

Predictor	OFF state	ON state	ΔG	Notes
ViennaRNA 1.8.5 (EteRNA)	-39.63	-35.50	+4.13	ViennaRNA 2.x says that the tandem UG is paired, while ViennaRNA 1.8.x says it's a loop. But still, the pairing probabilities are quite high, even for EteRNA.
ViennaRNA 2.1.x	-34.20	-32.50	+1.70

JR 4

Predictor	OFF state	ON state	ΔG	Notes
ViennaRNA 1.8.5 (EteRNA)	-38.13	-35.50	+2.63	ViennaRNA 2.1.x predicts that the ON state will be the MFE.
ViennaRNA 2.1.x	-30.20	-30.40	-0.20	ViennaRNA 2.1.x predicts that the ON state will be the MFE.

Salish - Salish's solve (removed)

Predictor	OFF state	ON state	ΔG	Notes
ViennaRNA 1.8.5 (EteRNA)	-30.53	-28.30	+2.23	VRNA 2.1.x predicts a misfold with the 5' side of the barcode hairpin in the OFF state.
ViennaRNA 2.1.x	-24.00	-23.80	+0.20

----

Designer's comments:

It was found that the code used here was an imperfect solve: Delta G is way too high, too many PlyA's, and a severe problem with the U in the 41 position, for mimics 3 (UGUAUU--GAAGG) and 4 (AGGAUA--UAAUU). Thus, the entrie series has been retracted.

Original code listed as a reference: GGAAAAAAAGUGCCAGGAUAUGGUUGUGUCAAAAAUUAAUUAGAAGGGGCACAACAAAAAAACAAAAAGUUUUUAAAGUAUAAAAAAACCCCUUCUUCGGAAGGGGAAAAGAAACAACAACAACAAC

Salish99 (talk) 08:41, 7 April 2014 (UTC)

Salish99 - Salish02

Predictor	OFF state	ON state	ΔG	Notes
ViennaRNA 1.8.5 (EteRNA)	-30.53	-30.40	+0.13	ViennaRNA 2.1.x predicts that the ON state will be the MFE.
ViennaRNA 2.1.x	-24.00	-25.90	-1.90	ViennaRNA 2.1.x predicts that the ON state will be the MFE.

Series dotplot: http://rna.tbi.univie.ac.at/RNAfold/ul0tLA9EX3/sequence1_dp.pdf

Salish99 - Salish Mark.3

Predictor	OFF state	ON state	ΔG	Notes
ViennaRNA 1.8.5 (EteRNA)	-29.23	-28.80	+0.43	ViennaRNA 2.1.x predicts that the ON state will be the MFE.
ViennaRNA 2.1.x	-24.10	-26.00	-1.90	ViennaRNA 2.1.x predicts that the ON state will be the MFE.

Series dotplot: http://rna.tbi.univie.ac.at/RNAfold/MznSrzLjIk/sequence1_dp.pdf

Hyphema - Tenacity

Predictor	OFF state	ON state	ΔG	Notes
ViennaRNA 1.8.5 (EteRNA)	-34.53	-33.90	+0.63	ViennaRNA 2.1.x predicts that the ON state will be the MFE.
ViennaRNA 2.1.x	-28.80	-30.30	-1.50	ViennaRNA 2.1.x predicts that the ON state will be the MFE.

Hyphema - Mod of JR Sub1

Predictor	OFF state	ON state	ΔG	Notes
ViennaRNA 1.8.5 (EteRNA)	-37.23	-35.60	+1.63	Additional 28-33 pair in the ON state for VRNA 1.8.5.
ViennaRNA 2.1.x	-30.50	-29.30	+1.20	Additional 28-33 pair in the ON state for VRNA 1.8.5.

Salish99 - Salish Var06

Predictor	OFF state	ON state	ΔG	Notes
ViennaRNA 1.8.5 (EteRNA)	-32.93	-32.50	+0.43	ViennaRNA 2.1.x predicts that the ON state will be the MFE.
ViennaRNA 2.1.x	-27.10	-29.00	-1.90	ViennaRNA 2.1.x predicts that the ON state will be the MFE.

----

Designer's comments:

Series dotplot: http://rna.tbi.univie.ac.at/RNAfold/9BqiQ3jOP6/sequence1_dp.pdf

@Salish99: -29.97 is the ensemble FE, which is something different than the MFE. And I would suggest to use http://rna.tbi.univie.ac.at/cgi-bin/RNAeval.cgi to obtain the FE of any structure. -- ElNando888 (talk) 13:17, 7 April 2014 (UTC)

Figured that was the summarized FE, but I couldn't figure out how to extract the structure denomination from the lab... How do I do that? It's automatically given in the challenges... Salish99 (talk) 19:32, 7 April 2014 (UTC)

The design button in the lab gives you something like http://eterna.cmu.edu/game/puzzle/4139129/
Just replace "game" by "web" in the URL, like this: http://eterna.cmu.edu/web/puzzle/4139129/
Yes, annoying that we have to do it by hand, but it works...
-- ElNando888 (talk) 20:25, 7 April 2014 (UTC)

Thanks for the tip. Tried it with http://eterna.cmu.edu/web/puzzle/4139477/ and this solution GGGAGGAAGGUGCGCCUAUCUGGUUGUGUGAAGAAUUAAUCAGAAGGCGCACAACAAUAAUCGGUUUACGAUUAGAAGUAUGAAUAAGCGCUGUCUUCGGACAGCGAAAAGAAACAACAACAACAAC

but: Sequence and structure are of unequal length!

I'm still doing something wrng here... 92.226.205.52 09:48, 8 April 2014 (UTC)

Ah, sorry, I'm so used to it that I forgot to mention: lab puzzles have implicit "tails", which means that you need to add 2 dots at the beginning, and 20 at the end. -- ElNando888 (talk) 11:04, 8 April 2014 (UTC)

Thanks. Works well with designs. But what about manufactured solutions?

Incidentally, I tried to use the http://171.64.65.150:8080/media/Score_Seq_SS.html to seee how the score coincides with my results, for example http://eterna.cmu.edu/game/browse/3553120/?filter1=Id&filter1_arg1=3725371&filter1_arg2=3725371

However, when replacing "game" with "web", I get this error:

/web/solution/.../seeresult

This does not seem to work for synthesized RNA's...

In the URL you gave, 3553120 is the puzzle, 3725371 is the specific design. So, I would use:

http://eterna.cmu.edu/web/puzzle/3553120/ to access the secondary structure, and
http://eterna.cmu.edu/game/solution/3553120/3725371/copyandview/ to get the design's sequence.

-- ElNando888 (talk) 19:55, 16 April 2014 (UTC)

ok, thanks

Salish99 (talk) 09:46, 8 December 2014 (UTC)

Hyphema - Reality

Predictor	OFF state	ON state	ΔG	Notes
ViennaRNA 1.8.5 (EteRNA)	-36.13	-35.50	+0.63	ViennaRNA 2.1.x predicts that the ON state will be the MFE.
ViennaRNA 2.1.x	-29.40	-31.60	-1.20	ViennaRNA 2.1.x predicts that the ON state will be the MFE.

ElNando888 - Meemeek

Predictor	OFF state	ON state	ΔG	Notes
ViennaRNA 1.8.5 (EteRNA)	-47.23	-42.20	+5.03	ViennaRNA 1.8.5 predicts this sequence cannot switch. Both states feature non-canonical pairs.
ViennaRNA 2.1.x	-40.10	-37.20	+2.90
ViennaRNA (Andronescu 07)	-38.60	-37.49	+1.11

Brourd - Design 1

Predictor	OFF state	ON state	ΔG	Notes
ViennaRNA 1.8.5 (EteRNA)	-34.83	-30.30	+4.53	ON state features a non-canonical pair.
ViennaRNA 2.1.x	-27.80	-25.80	+2.00	ON state features a non-canonical pair.

ElNando888 - Meemeeq

Predictor	OFF state	ON state	ΔG	Notes
ViennaRNA 1.8.5 (EteRNA)	-40.13	-39.60	+0.53	ViennaRNA 2.1.x predicts that the ON state will be the MFE. Both states feature non-canonical pairs.
ViennaRNA 2.1.x	-32.50	-35.40	-2.90
ViennaRNA (Andronescu 07)	-34.76	-33.63	+1.13

Iaxtrdinar - Rx for FMN mimic

Predictor

OFF state

ON state

ΔG

Notes

ViennaRNA 1.8.5 (EteRNA)

-33.00

-31.40

+1.60

OFF state has additional pairs in the large apical loop in VRNA 1.8, and features a non-canonical pair.

ViennaRNA 2.1.x predicts that the ON state will be the MFE.

ViennaRNA 2.1.x

-24.90

-28.40

-3.50

Jnicol - Polly 1

Predictor	OFF state	ON state	ΔG	Notes
ViennaRNA 1.8.5 (EteRNA)	-44.73	-40.70	+4.03	.
ViennaRNA 2.1.x	-38.60	-37.10	+1.50	.

Oolong 1

Predictor	OFF state	ON state	ΔG	Notes
ViennaRNA 1.8.5 (EteRNA)	-26.83	-26.10	+0.73	The OFF state is predicted to be missing the entire neck stack by ViennaRNA 2.1.x.
ViennaRNA 2.1.x	-21.80	-20.80	+1.00

Oolong 2

Predictor	OFF state	ON state	ΔG	Notes
ViennaRNA 1.8.5 (EteRNA)	-29.53	-25.00	+4.53	ViennaRNA 2.x predicts a misfold of the barcode in the unbound version.
ViennaRNA 2.1.x	-26.50	-24.20	+2.30

Oolong 3

Predictor	OFF state	ON state	ΔG	Notes
ViennaRNA 1.8.5 (EteRNA)	-33.53	-28.90	+4.63	.
ViennaRNA 2.1.x	-29.90	-25.40	+4.50	.

Jnicol - Polly 2

Predictor	OFF state	ON state	ΔG	Notes
ViennaRNA 1.8.5 (EteRNA)	-39.83	-38.80	+1.03	ViennaRNA 2.1.x predicts that the ON state will be the MFE.
ViennaRNA 2.1.x	-33.80	-35.20	-1.40	ViennaRNA 2.1.x predicts that the ON state will be the MFE.

Eli Fisker - D1

Predictor	OFF state	ON state	ΔG	Notes
ViennaRNA 1.8.5 (EteRNA)	-31.53	-29.80	+2.73	ViennaRNA 2.1.x predicts that the ON state will be the MFE.
ViennaRNA 2.1.x	-27.30	-27.90	-0.60	ViennaRNA 2.1.x predicts that the ON state will be the MFE.

Eli Fisker - D2

Predictor	OFF state	ON state	ΔG	Notes
ViennaRNA 1.8.5 (EteRNA)	-30.73	-29.40	+1.33	ViennaRNA 2.1.x predicts that the ON state will be the MFE.
ViennaRNA 2.1.x	-26.30	-26.40	-0.10	ViennaRNA 2.1.x predicts that the ON state will be the MFE.

Eli Fisker - D3

Predictor	OFF state	ON state	ΔG	Notes
ViennaRNA 1.8.5 (EteRNA)	-31.63	-30.50	+1.13	ViennaRNA 2.1.x predicts that the ON state will be the MFE.
ViennaRNA 2.1.x	-27.40	-28.70	-1.30	ViennaRNA 2.1.x predicts that the ON state will be the MFE.

Jieux - Mimitch A

Predictor	OFF state	ON state	ΔG	Notes
ViennaRNA 1.8.5 (EteRNA)	-41.03	-38.50	+1.53	.
ViennaRNA 2.1.x	-33.70	-33.50	+0.20	.

Jieux - Mimitch B

Predictor	OFF state	ON state	ΔG	Notes
ViennaRNA 1.8.5 (EteRNA)	-38.13	-35.90	+2.23	ViennaRNA 2.1.x predicts that the ON state will be the MFE.
ViennaRNA 2.1.x	-29.70	-30.90	-1.20	ViennaRNA 2.1.x predicts that the ON state will be the MFE.

Jieux - Mimitch C

Predictor	OFF state	ON state	ΔG	Notes
ViennaRNA 1.8.5 (EteRNA)	-40.23	-38.90	+1.33	ViennaRNA 2.1.x predicts that the ON state will be the MFE.
ViennaRNA 2.1.x	-32.70	-34.60	-1.90	ViennaRNA 2.1.x predicts that the ON state will be the MFE.

mat747 - Slot Filler 1

Predictor	OFF state	ON state	ΔG	Notes
ViennaRNA 1.8.5 (EteRNA)	-31.53	-29.80	+1.73	ViennaRNA 2.1.x predicts that the ON state will be the MFE.
ViennaRNA 2.1.x	-27.30	-27.90	-0.60	ViennaRNA 2.1.x predicts that the ON state will be the MFE.

mat747 - Slot Filler 2

Predictor	OFF state	ON state	ΔG	Notes
ViennaRNA 1.8.5 (EteRNA)	-31.03	-29.90	+1.13	ViennaRNA 2.1.x predicts that the ON state will be the MFE.
ViennaRNA 2.1.x	-26.80	-28.10	-1.30	ViennaRNA 2.1.x predicts that the ON state will be the MFE.

mat747 - Slot Filler 3

Predictor	OFF state	ON state	ΔG	Notes
ViennaRNA 1.8.5 (EteRNA)	-31.33	-30.00	+1.33	ViennaRNA 2.1.x predicts that the ON state will be the MFE.
ViennaRNA 2.1.x	-27.00	-28.00	-1.00	ViennaRNA 2.1.x predicts that the ON state will be the MFE.

Brourd - Design 2

Predictor	OFF state	ON state	ΔG	Notes
ViennaRNA 1.8.5 (EteRNA)				.
ViennaRNA 2.1.x				.

Brourd - ThreDesiBrourdegn

Predictor	OFF state	ON state	ΔG	Notes
ViennaRNA 1.8.5 (EteRNA)				.
ViennaRNA 2.1.x				.

.