puzzle picture
1964: Unsolved Huge Cryo-EM Density: Beet Virus
Status: Closed

Summary

Name: 1964: Unsolved Huge Cryo-EM Density: Beet Virus
Status: Closed
Created: 03/03/2021
Points: 100
Expired: 03/18/2021 - 18:00
Difficulty: Expert
Description: We are very excited about this puzzle, because our collaborators who provided us with the 4 structures from Foldit's 2019 cryo-EM paper have been unable to solve this cryo-EM structure. This will not be easy, because the full density has 46 symmetric monomers and we can't cut out a single copy of density for you (since we don't know the structure). Instead, we've sliced off a quarter of the entire density, which our collaborators estimate should be ~10 monomers. So keep this in mind, as we are only providing you with a single monomeric chain to fold. More details and pictures in the puzzle comments!
Categories: Electron Density, Overall

Top Groups

RankGroupScorePoints
6Gargleblasters14,37914
7Contenders14,2638
8Void Crushers13,0285
9Olsztynek13,0003
10Australia12,7422

Top Evolvers

Top Soloists



Need this puzzle? Log in to download.  

Comments

beta_helix's picture
User offline. Last seen 16 hours 51 min ago. Offline
Joined: 05/09/2008
Groups: None
More details for this difficult puzzle

This 3.5Å cryo-EM map is the nucleocapsid of a plant virus and contains a long RNA molecule that follows the helical symmetry.
Therefore, some of the electron density that you have been given should not be filled with protein.

Sequence:
MSSEGRYMTWKDMSHNKFMTDRWARVSDVVSVIKQSHAMDLSKAANLS
IIKTALAGLGSGWSDSNPFVSPMTRFPQTLTMYGALVLYVNLSDPEFA
LIMTKVNTLTDSGLADNASANVRRDVVSGNKAESSGKTAGTNENSAYT
LTVSLAGLAQALRLEELMWTRDKFEDRFKLPWTPVQGRTSPPGQ

The full cryo-EM map has 15.25 monomers per helix turn (side-view shown in image below).
We didn't want to provide you with this full map, as it fits ~46 monomers, so instead we have given you a large "slice" that should contain around 10 monomers (the blue "donut slice" in the top-view image at the very bottom).

We obviously do not know the precise details, because the only way to perfectly trim the cryo-EM map down to a single chain would require solving the structure! We know this puzzle will not be easy, which is why we are providing you with 14 days to work on it... good luck!


Susume's picture
User offline. Last seen 6 days 16 hours ago. Offline
Joined: 10/02/2011
All the same pose?

Are all the monomers assumed to have the same shape and a similar place in the donut? Like if one of them contacts the inside of the donut, then all of them contact the inside of the donut in the same way?

beta_helix's picture
User offline. Last seen 16 hours 51 min ago. Offline
Joined: 05/09/2008
Groups: None
Great question!

From what we know about this protein: most likely yes, they are all the same pose.

jeff101's picture
User offline. Last seen 2 hours 26 min ago. Offline
Joined: 04/20/2012
Groups: Go Science
Does the ED cloud contain RNA too?
"This 3.5Å cryo-EM map is the nucleocapsid of a plant virus and 
contains a long RNA molecule that follows the helical symmetry.
Therefore, some of the electron density that you have been given 
should not be filled with protein."

Does this mean some of this puzzle's ED cloud is due to RNA?
Should we expect to see odd non-protein ED in its cloud?
beta_helix's picture
User offline. Last seen 16 hours 51 min ago. Offline
Joined: 05/09/2008
Groups: None
Yep!

That is indeed what we expect.

LociOiling's picture
User offline. Last seen 7 hours 44 min ago. Offline
Joined: 12/27/2012
Groups: Beta Folders
JPred results, etc.

Jpred was in a good mood today, so here's the results page for this protein: puzzle 1964 prediction

That link will probably expire at some point, I think the predictions have only a finite shelf life.

The first match is to a sequence known as UniRef90_G1BIV2, which gets you to a paper Beet soil-borne mosaic virus RNA-3 is replicated and Beet soil-borne mosaic virus RNA-3 is replicated and encapsidated in the presence of BNYVV RNA-1 and -2 and allows long distance movement in Beta macrocarpa in the presence of BNYVV RNA-1 and -2 and allows long distance movement in Beta macrocarpa from 2009.

My main takeaways so far are 1) "encapsidated" is a word, 2) Beta macrocarpa is a scientific name for a type of beet, and 3) that paper has a really long title.

(Maybe we'll change our group name to "Beet Folders".)

beta_helix's picture
User offline. Last seen 16 hours 51 min ago. Offline
Joined: 05/09/2008
Groups: None
Hahahahahahahahaha

Thanks LociOiling,

If your team solves this structure, the Beta Folders should totally change their name to Beet Folders! :-P

frood66's picture
User offline. Last seen 1 hour 40 min ago. Offline
Joined: 09/20/2011
Groups: Marvin's bunch
just an aside

do u guys at Foldit think this ed will actually help? or are u just looking for a lucky pose.?

no I am not being rude....I simply do not understand the point of this puzzle - nor the use of an ed.

beta_helix's picture
User offline. Last seen 16 hours 51 min ago. Offline
Joined: 05/09/2008
Groups: None
Lucky pose would need to be very lucky in this case!

I believe that this ED will help, but indeed it will not be easy (hence the Expert difficulty).

Since the ED cloud boosts your Foldit score if it matches your model, you can think of it as:
The highest possible Foldit score on this puzzle if you ignore the ED is some number X,
but the highest possible Foldit score on this puzzle (if you are able to match the ED correctly) is going to be a lot higher than X.

Susume's picture
User offline. Last seen 6 days 16 hours ago. Offline
Joined: 10/02/2011
Similar virus to view structure

This is not the same virus (not even similar sequence I don't think), just one that has similar architecture to help folks view it.
https://www.rcsb.org/3d-view/1CGM
Like our puzzle, this virus forms a hollow cylinder with about 15-16 copies of the protein making one full turn, then continues into another turn - it's a very flat helix that ends up looking kind of like a layer cake. The protein's job is to pack around the RNA to protect it - you can see the RNA (on my screen it's green) tucked in between the protein layers.

Every copy of the protein has to touch the inner surface of the donut - which means it's pretty crowded in there! There's only room for each protein to form one loop in the center, then it has to head back out toward the outer surface. If you can find the identical loops on the inner surface in our puzzle ED, you can mark them with colored dots to help you decide which protein copy to follow (I used red for all the ones I didn't want and green for the one I picked).

Similarly, every copy has to take up about 1/16 of the outer donut surface in its 'cake layer', so you can look for repeated shapes in the outer layer to mark. Finally, if you look straight down through the donut in this sample virus, you can see that the 'wedges' formed by the individual proteins are not straight-sided triangles but sort of crescent-shaped - I don't know if that's true in our puzzle, but it's something to look for.

jeff101's picture
User offline. Last seen 2 hours 26 min ago. Offline
Joined: 04/20/2012
Groups: Go Science
For Iron Maiden fans:

1964, the Number of the Beet.

jeff101's picture
User offline. Last seen 2 hours 26 min ago. Offline
Joined: 04/20/2012
Groups: Go Science
For fans of the Go-Go's:

We Got The Beet.

jeff101's picture
User offline. Last seen 2 hours 26 min ago. Offline
Joined: 04/20/2012
Groups: Go Science
For fans of Wendy's:

Where's The Beet?

Susume's picture
User offline. Last seen 6 days 16 hours ago. Offline
Joined: 10/02/2011
Top and bottom of cloud cut off too?

The cloud appears complete on the inner and outer donut surfaces, and you've told us it has been sliced through the donut to make a wedge. What about the other two planes, the ones parallel to the plate if you set the donut on a plate? Are they sliced through density that extends further in that direction, or is our cloud as tall as the donut gets?

beta_helix's picture
User offline. Last seen 16 hours 51 min ago. Offline
Joined: 05/09/2008
Groups: None
These are all the other images that our collaborators sent us:

Based on the image above, it definitely seems like the density has been sliced across the top and underneath it as well. I can ask the experimentalists if you like.

Susume's picture
User offline. Last seen 6 days 16 hours ago. Offline
Joined: 10/02/2011
seems clear

That top one does look like it's been sliced top and bottom. Thanks!

Joined: 09/24/2012
Groups: Go Science
I don't undestand what to do with this puzzle

Thanks for your comments but I'm still in the cloud. In the top image (similar with the ED we received), it's like a repetition 3.5 x the same cloud. Should we try to fit the protein into one of them (say in the middle)?

If it's th case, have we a tool to slice the density map in order to only keep what we want to use ?

beta_helix's picture
User offline. Last seen 16 hours 51 min ago. Offline
Joined: 05/09/2008
Groups: None
Can you make it to office hours on 3/11?

Hi Bruno,
I plan to discuss all this at tomorrow's office hours:
https://fold.it/portal/node/2011345

You are correct that you just need to fit the protein into one "slot" of density, keeping in mind that RNA fits somewhere in there as well.

In order to trim the density, you can use the Trim Density tool:
https://fold.it/portal/node/995580 (scroll down a bit)

horowsah's picture
User offline. Last seen 6 days 14 hours ago. Offline
Joined: 03/17/2016
Groups: DU Proteins S20
radius view

This may or may not address some of the issues with the large amount of density, but if you haven't tried using the radius slider yet, I think it helps for this sort of challenge.

APPAAP's picture
User offline. Last seen 4 weeks 6 days ago. Offline
Joined: 04/25/2020
Groups: None
Cryo-EM Density Given

I think that the Cryo-EM Density given is the 1/8th of the whole nucleocapsid of the plant virus. Clearly there is the inside and outside of the cross section donut shape shown above. There are also 4 cut planes say the TOP, BOTTOM, LEFT and RIGHT, and I think that the the TOP or BOTTOM planes is somewhere in the middle of the nucleocapsid as shown in the whole donut cross section the other one TOP or BOTTOM is indeed the TOP or BOTTOM of Cryo-EM density of the nucleocapsid. Do you agree or I am missing something?

beta_helix's picture
User offline. Last seen 16 hours 51 min ago. Offline
Joined: 05/09/2008
Groups: None
1/4 instead of 1/8th?

Looking at the blue cross-section in the first post: https://fold.it/portal/files/images/BNYVV_map-both.png
I believe we are providing you with 1/4th of the whole density.

We can discuss this more in 30 minutes at the Foldit Office Hour about this.

Susume's picture
User offline. Last seen 6 days 16 hours ago. Offline
Joined: 10/02/2011
less than a quarter

I think AAPAAP may be referring to the fact that if you have a quarter donut, and you shave off the top quarter and bottom quarter of that piece, you will be left with an eighth of a donut.

By the way, even though it would make the cloud bigger, I would prefer the top and bottom not be sliced off, because they would help define what makes a monomer unit. I have one helix that could go either on the top or the bottom of the other ones - if I could see the top and bottom of the donut, I would know which side it goes on.

jeff101's picture
User offline. Last seen 2 hours 26 min ago. Offline
Joined: 04/20/2012
Groups: Go Science
Some Questions:

I missed today's office hour, so below are some questions I have:

(1) In https://fold.it/portal/node/2011305#comment-44223
where is the RNA? Is it inside the empty hole at the center
of the donut, so in effect all of the ED cloud we are given
is from protein? Is the RNA mixed with protein in the ED
cloud we are given? What does the ED typically look like 
for RNA? Long ago you gave us images of the ED shapes for 
different protein sidechains. Are there similar pictures
you could give us for pieces of RNA?

(2) When the protein interacts with RNA, should we expect
hydrophobic or hydrophilic residues at the protein/RNA
interface? Does this mean the inner surface of the donut
should be full of hydrophobic residues? How about the 
outer surface of the donut? Are there certain protein
sidechains or sequences that often bind to RNA?

jeff101's picture
User offline. Last seen 2 hours 26 min ago. Offline
Joined: 04/20/2012
Groups: Go Science
Susume's reference:
https://www.rcsb.org/3d-view/1CGM gives a 3D image of 
a donut structure that you can move and rotate on the 
screen. I think the brown part is protein and the green 
part is RNA. In the brown part, each protein subunit has 
both its termini at the outer circle of the donut. The 
green part (RNA?) is close to the inner circle of the 
donut, but some brown parts are even closer to the inner 
circle. 

https://www.rcsb.org/structure/1cgm has 2 images:
(1) "Biological Assembly 1" looks like our donut's ED. 
You can download this image as 1cgm_assembly-1.jpeg.
(2) "Asymmetric Unit" gives 1cgm_model-1.jpeg, which
has a brown part that looks like protein and a green
part that could be RNA. If you click on the right 
spot, it goes to the link below, which is a rotatable
3D image of a single protein/RNA subunit:
https://www.rcsb.org/3d-view/1CGM/0
Overall it seems like 1CGM has 4 helices and many
long loops per subunit.
Susume's picture
User offline. Last seen 6 days 16 hours ago. Offline
Joined: 10/02/2011
not our virus

Keep in mind 1CGM is not our same virus - I just typed in 'mosaic virus' because I knew they had a similar capsid structure and took the first one that looked good. The psipred predicted secondary structure that came with the puzzle is more likely to be accurate than trying to match the secondary structure of 1CGM.

Joined: 09/29/2016
Groups: Gargleblasters
For what its worth...

My trROSETTA results (from YangLab's webserver) used this as its top match when coming up with its end result:
https://www.rcsb.org/3d-view/ngl/3pdm
Though, the 1CGM was its second match.

My 'tr' result ended up being similar, but still having a few differences.

I ahdn't realized that the mega-structure these formed was made up of so many of these monomers, so today when I went in and mapped where I thought each monomer was, I was coming up with only 3 or so... After seeing that one I now realize there's roughly a dozen in our section of cloud... DOH!
Back to the drawing board tomorrow! :(

CryoEMguy's picture
User offline. Last seen 13 weeks 2 days ago. Offline
Joined: 03/14/2021
Groups: None
Some replies

Hi Jeff,
here some more info :
(1) RNA is somewhere in the ED cloud (see my answer to another question below). About its shape : you have a phosphate-sugar backbone that will form some sort of "necklace", from which the bases will stick out (and most likely form short stacks). The bases look like large amino acids (roughly the size of a tyrosine).
(2) You can expect different types of interactions. The phosphate groups of the RNA (which,by the way, will be strong densities) can interact with positively charged amino acids like arginines or lysines. The bases may, in some cases, interact with hydrophobic amino acids : you can even have stacking of bases to aromatic residues such as phenylalanine.

jeff101's picture
User offline. Last seen 2 hours 26 min ago. Offline
Joined: 04/20/2012
Groups: Go Science
Are all RNA subunits in 1964 the same?

If Puzzle 1964's structure is like 1CGM's, 
most of the ED is due to protein, and only
a single RNA subunit (A C G or U) is 
included for each protein subunit. Should
we expect all of the RNA subunits to be
identical (like an RNA sequence of all A's),
or should we expect there to be a mix of
A C G & U in the ED? In 1964, should we 
expect the overall RNA to be one long strand, 
or should we expect there to be base-pairing 
(A-U and C-G) between 2 long strands of RNA?

Thanks again,
Jeff

CryoEMguy's picture
User offline. Last seen 13 weeks 2 days ago. Offline
Joined: 03/14/2021
Groups: None
Are all RNA subunits in 1964 the same?

Hi Jeff,
yes most of the density is due to protein, there are only a few nucleotides per protein subunits, probably between 3-5 (but let's consider this to be an unknown). All the RNA bases will appear identical, actually they are averaged out by the image processing, and therefore represent an average of all possible bases present in the viral genome. The RNA backbone should be clearly identifiable.
The overall RNA should be one long, single strand (so, no base pairing).
Hope that helps
A.

beta_helix's picture
User offline. Last seen 16 hours 51 min ago. Offline
Joined: 05/09/2008
Groups: None
Thank you, A!

CryoEMguy is our collaborator who sent us the structure and density for this puzzle :-)

APPAAP's picture
User offline. Last seen 4 weeks 6 days ago. Offline
Joined: 04/25/2020
Groups: None
Up correction

To Susume up

Thank you Susume, that I mean, I cannot see clearly from cryo-EM map provided where the top or bottom of the nucleocapsid of a plant virus is, even thought I can identify part of the RNA helix. Also my big problem is how to fold the protein (I think trial and error) Jpred prediction above shows some sheet structure, are you using only α-helices and loops or/and β-sheet structures?
At the end I do not know how to fix almost known points of the protein to almost known points of the cryo-EM map. The last need some strategy and practice with the Foldit program cryo-EM map density.

jeff101's picture
User offline. Last seen 2 hours 26 min ago. Offline
Joined: 04/20/2012
Groups: Go Science
ED Trim Density Tool
In Puzzle 1964, I have found the Trim Density tool useful.
It seems very sensitive about 1mm from the Far end, and
that is the range I've found to work best. Having this
slider report a number for its present setting (and perhaps 
giving a finer resolution of settings about 1mm from the 
Far end) would make this slider more useful.
jeff101's picture
User offline. Last seen 2 hours 26 min ago. Offline
Joined: 04/20/2012
Groups: Go Science
Make it easier to remove dots

On this puzzle, I've put so many dots in the ED cloud
that they are no longer helpful. I want to delete some
but not all of them. Removing dots by hand is pretty
tedious right now. It seems like I have to Tab on a dot
and then move the mouse to click on a Trash Can icon
that appears. It would be better if a single button 
like Delete or Backspace instead of Tab could be 
pressed, and this would remove a dot.

Also, Puzzle 1964 has many of the same problems that
Puzzle 1667: "Too Much Density!" Freestyle puzzle 
(https://fold.it/portal/node/2007726) had, and
https://fold.it/portal/node/2007742 lists many ideas 
to improve ED puzzles based on problems with Puzzle 
1667. If ED puzzles are to become more frequent in 
Foldit, please implement some of the ideas in the 
above Feedback.

One quirky thing I should mention is that when I
added a label to a dot in the Selection Interface,
I had to Tab on this dot to see its label. Also,
Note Mode would show this label in the Original 
Interface but not in the Selection Interface.

Joined: 09/29/2016
Groups: Gargleblasters
Wholly agree!

The need for a numerical value on the Trim slider would be helpful.
Also, that it be explained somewhere (or maybe it is, these days) on how that Trim tool works. Not so much in the overall function, but that there are factors and quirks (it seems).

For example, that when using the Trim menu, it will always reset the slider to its full amount. Therefore, if you Trim, do stuff, and open the Trim menu again... if you just click OK, it removes your Trimmed amount. If you slide it down some, but to a value that was not as low as what you used before, when you click OK that will add some density back!
So, even if there's a number value included... either the slider needs to actively remember what your slider was previously set at, or you need to have it at least display the previous value that was used, so we have that as a reference.
(For months I had thought the Trim tool's slider worked in an "Additive" way, where by sliding it down 3 notches, then coming back and sliding it down 3 more notches, resulted in a total Trimming of ""6"". That's usually how software has worked in my experience, where a slider that resets conveys a sense of "Ok, your previous value is now the new baseline (max slider value), how much change would you like to do this time". Wasn't until I needed to reset the cloud and just decided to try it at Max Slider, did I realize that's how to reset it lol Also that I had been not trimming my density how I had wanted to, that entire time!)

The other thing is that the Protein's shape apparently is what the Trim tool is working off of... Which I only found out in this puzzle :S
I wanted to try and trim out juuust a single Monomer's density, and hadn't yet folded the protein, so I Centered Protein on Density and ran the Trim, expecting it to sort of work like a combination of Radius and Threshold (admittedly, I don't know why I expected that it'd work that way... lol), yet I was left with a tinny little cylinder of Density around my toothpick protein. Doh! :P

At any rate, all three of Jeff's suggestions (though the Density Notes might just be a bug) above get a thumbs-up from me!

APPAAP's picture
User offline. Last seen 4 weeks 6 days ago. Offline
Joined: 04/25/2020
Groups: None
ED trim

I realised that the ED given is very close to 1/4 or just 1/4, may be on top or bottom there is no cut planes.
The monomer (protein given) can only be seen fully in the ED given in the middle (there are almost 12 monomers but not all are fully seen in the ED given) because at the ends there are at least two cut planes. The RNA is best seen from top and closer to the inside part than to outside part.
Although this seems simple, I find extremely difficult to align. My protein fold is completely out of phase with the ED given. I need to refold but may be is not good practice refold only to achieve a higher score. I must learn using the ED trim tool with simpler cases.

frood66's picture
User offline. Last seen 1 hour 40 min ago. Offline
Joined: 09/20/2011
Groups: Marvin's bunch
the only prob

this puzzle flips out too often.....please can that be sorted?. is it that u guys are asking too much? or is it that we have machines too small for it?

maybe now would be a good time for us to be advised about minimum machine size....internet speed etc? it would save a lot of heartache.

just saying

jeff101's picture
User offline. Last seen 2 hours 26 min ago. Offline
Joined: 04/20/2012
Groups: Go Science
1964 has crashed many times in my hands

Seems like when I try to turn the ED on while a recipe is running, 
1964 crashes. Also, if I am doing shake & trying to add a band at 
the same time, I often get a crash. Sometimes when I add bands,
the response is slow, and if I move the mouse too soon, it moves
my band from where I wanted it to be. Many times the Foldit
window has said "not responding", and if I try to do another
action, it crashes. When I see the message, it seems better to
wait for the message to go away before trying my next action.

1964 has generally been well-behaved when running recipes
(if I keep the ED off or don't try to change the ED view
settings).

beta_helix's picture
User offline. Last seen 16 hours 51 min ago. Offline
Joined: 05/09/2008
Groups: None
Thanks for the details

Thank you for these examples, as we can hopefully now reproduce these crashes on our end.

We realize a puzzle with density this big is pushing the limits of what Foldit and your computers can handle, but that doesn't mean it should be crashing this badly.

Now that we have your steps to reproduce it, we're looking into a solution and hope to resolve these crashes soon.
Thanks for you patience!

jeff101's picture
User offline. Last seen 2 hours 26 min ago. Offline
Joined: 04/20/2012
Groups: Go Science
Viewing the ED in each unit cell of the protein complex:
I think for Puzzle 1964's helical protein/RNA complex,
you could break up the overall structure into many 
identically-shaped pieces, a bit like pie pieces in 
the board game Trivial Pursuit. The tops and bottoms 
of these pieces would be slightly-slanted so that if
you made a long enough ring of them, they would not
form a perfect circle. Instead, there would be almost
circular layers one on top of another, forming an 
overall helical shape. Each of these pieces would 
contain all the segments for a complete protein 
monomer, but it might be segments 1-30 for one 
monomer, segments 31-120 for a second monomer, and 
segments 121-188 for a third monomer, for example.

It would be neat if we could see the ED for just 
one of these pieces, and if our protein crossed 
through any of the boundaries for this piece, it
would get chopped off at that boundary. The rest
of the protein would continue through the correct
point on another part of the piece's boundaries, 
and it would continue within the same piece of 
the ED. This way, if our protein straddled several
of the repeats within the overall structure, we
could easily check that it wasn't overlapping 
with another copy of itself. Perhaps the Foldit
score could account for interactions of proteins 
with other copies of themselves and help prevent
them from overlapping with each other.

Having a small chunk of the ED like above would 
make it easier to spot features within the ED
and mark them with dots.

https://fold.it/portal/node/2007742#comment-38508
discusses similar things for a simpler complex
that has box-shaped repeats, more like the 
typical unit cell discussed in chemistry.

Finally, if you can break up the experimental 
ED into many repeat units, maybe you can average 
the ED from many of these different repeat units 
to give an ideal repeat unit's ED with better 
signal to noise.
frood66's picture
User offline. Last seen 1 hour 40 min ago. Offline
Joined: 09/20/2011
Groups: Marvin's bunch
with regard to viewing options

I have long asked for the ability to erase parts/volumes of the ED that I do not want to see.

sure...we have trim to density, radius and a multitude of viewing options. But the truth is...the more we put in the cloud, the harder it gets to look at what we want to.

how about a selection option....where only the selected segments count....all others disappear along with any ED associated with them.

Is this worth a punt?

anything to avoid headaches and sore eyes imao

User login
Download links:
  Windows    OSX    Linux  
Windows
(7/8/10)
OSX
(10.12 or later)
Linux
(64-bit)

Are you new to Foldit? Click here.

Are you a student? Click here.

Are you an educator? Click here.
Social Media


Search
Only search fold.it
Other Games: Mozak
Recommend Foldit
Topics
Top New Users
Sitemap

Developed by: UW Center for Game Science, UW Institute for Protein Design, Northeastern University, Vanderbilt University Meiler Lab, UC Davis
Supported by: DARPA, NSF, NIH, HHMI, Amazon, Microsoft, Adobe, RosettaCommons