Fight the COVID-19 with Eterna

This thread is to solicit general feedback about the scientific challenges and funding campaigns underway to fight COVID-19. As specific topics become popular, let’s create separate forum posts to organize discussions and link out from this pinned central forum post.

Please see the news post for more info:


Virus Hunter - or the story about how vaccine making was started

Should someone be interested in historic science writing, I can recommend the book Virus Hunter by Greer Williams. 

It is the story about how pioneers figured out first what was the viruses that caused disease and then their work in finding a vaccine. 

There are lots of diseases we are no longer worried about thanks to vaccines. Regularly history is helpful for putting current problems in perspective. 

The first vaccine was made by Dr. Jenner. Second chapter in part one is called "The man who listened to a Milkmaid.

It was a milkmaid who made the observation that lead to the pock vaccination. Those days there was nothing called citizen science. But that was what was. :slight_smile:

I have written a bit more about the book in this previous forum post. Check second part of the post. 

Our viral descent


Dev chat from today:

One of the things mentioned in the chat was the idea of an energy model sandbox, proposed by wayment.  This looks similar to an idea which I have thought about, that of allowing scoring scripts to be written.  When players solve puzzles there is a criterion of success.   For the labs this is absent, which can be a problem for new players - it is very difficult to judge what is needed for success.
My idea would be for experienced players to be able to attach scoring scripts to the labs, which gave players a score, which they could then compete to maximize (like in the eternabot puzzles).  Suppose that there were 5 scripts A, B, C, D and E.  The best scoring solutions for each of them would be tested by experiment in the labs.  The criteria for success in the labs would then be used to judge which scoring function is best, giving a basis for improved scripts in a later round.  For instance the best solution might be one that scored well with both scripts B and E, indicating that a script combining the two is the way forward.

1 Like

Genome Wrapping and the coding level above viral RNA

Here is my quest to understand the coronavirus. I will sum up the story I began with my coronavirus puzzles plus add a lot more. I have asked Rhiju a bunch of questions and brought back the spoils. 

Coronavirus - Solenoid

Coronavirus is an RNA virus.

Image of the coronavirus from WHO:

What is COVID-19 All You Need To Know “Corona Virus”

Cynwulf28 puzzle Solenoid as when screenshot halfway in between target and native looks just like the covid19 when bound up inside the viral capsid as in WHO’s image.

Eli: I wonder if these nucleoproteins are binding anywhere specific sequence wise in the RNA? Plus if the RNA is really like a flat circle in a 3D ball of viral protein?

Rhiju: I haven’t been able to find good studies on what RNA motifs the nucleocapsid (‘N protein’) binds to perhaps that something players could help figure out!

eli: Yes. I recall reading about it earlier somewhere in relation to another virus

rhiju: yes, even in ebola it’s not clear how the RNA wraps around (or inside) the nucleocapsid protein filament. Fascinating question.

eli: Oh. It kind of also reminds me of tobacco mosaic virus protein capsid. While that is a spirally wrapping around the RNA. (In a tube) And the above is a 3D wrapping. (Sphere)

Coronavirus - From solenoid to yarn ball

I was wondering about if the corona virus really was folded up like a flat circle (2D) in a ball of viral protein (3D) as in the WHO image.

Molecular Landscapes by David S. Goodsell

I was really happy when I found David Goodsell’s Covid-19 image. His image gave me an answer to what I was wondering about in relation to how the RNA was packed up inside the viral capsid. Because he always makes his images based on views on actual molecules. (The article linked under the picture holds reference to the papers from which it is based on.) So the RNA strand both touches up with the capsid, but the RNA strands also line up against each other when more inside the capsid.

The viral RNA is much more bundled up as a yarn ball. With the one RNA curl lining up to the next one and parts of them touching the viral shell. Just like yarn strings lying close to each other, in a yarn ball. 

In a ball of yarn, one can get action if one grabs the outer tail. Whereas threads in lower layers are not available for interaction. 

eli: I wonder, which of the tails of the virus is the outer tail in the RNA yarn ball? The 5’ or the 3’? Or both, like in a yarn ball?

rhiju: Yea, I’ve been wondering that too! There’s some evidence that the 5’ and 3’ ends interact at some point in the virion life cycle, maybe throughout.

eli: Lol and sigh. Viral circle of doom. 

rhiju: check these papers out:

Sequence Motifs Involved in the Regulation of Discontinuous Coronavirus Subgenomic RNA Synthesis

Characterization of the RNA Components of a Putative Molecular Switch in the 3′ Untranslated Region of the Murine Coronavirus Genome

What is happening - RNA histones?

What you didn’t know about your RNA, by DNews

I kind of get the nucleoproteins for the corona RNA. While it does have a function of attracting and keeping the viral capsid in place, it is also kind of RNA histones. :slight_smile:

Nucleoprotein wrapping around the corona RNA:

Molecular Landscapes by David S. Goodsell

I think I have never before seen RNA bundled up in like sized and similar threads.

A viral chromosome/plasmid :wink:

Rhiju: “RNA histones”. Yes!!

eli: Lol. Is there really such a thing?

My search engine suggest I search for “DNA histones” when I try “RNA histones”

rhiju: [regarding RNA histones, check out this paper:

RNA epigenetics - A layer above RNA

Eli: That is amazing. So nucleoprotein positions determine the splicing positions, depending on the sequence like UUUUU in mRNAs.

I have been searching on and found other examples and I’m now thinking that RNA just like DNA can have epigenetics.

Epigenetics - An introduction By Armando Hasudungan

Just one example. Viral RNA can have methylations just like DNA. And it can matter to how the mRNA gets read by the ribosome. Plus how well it is translated.

Which makes me wonder if we can methylate corona mRNA to hell?

Here is a paper on viral epigenetics. It is most about DNA epigenetics, but it is starting to mention the phenomenon in RNA also: I only have access to the abstract.

@rhiju, another thing that makes me think that the nucleoproteins in coronavirus has a role in the very neat wound up RNA, is that I have read in Wikipedia under coronavirus, that coronaviruses have the longest viral RNA genomes. (26 to 32 kilobaser, or 26000 to 32000 bases)

While I’m aware that the viral capsid proteins have a role in packing the RNA tight, I just can’t imagine a yarn ball getting made by yarn being wound from the outside and in.

Hence I think that the nucleoproteins play a role in the packaging.

rhiju:  there is a preprint on SARS-CoV-2 methylations — let me dig it out

eli:  Thx for the corona methylation paper! Lol on this phrase “DNA nanoball”. I say RNA nanoball :wink:

@rhiju, if these “RNA histones” are different from human histones, then they may be useful as a target, to prevent the RNA from assembling in a ball. From what I understand histones don’t mutate a lot, so it may be a more stable target to attack.

The paper was a most interesting read. Especially cool that they found a connection with the most chemically modified mRNA to have the shortest polyA tails.

Afterthought, perhaps corona RNA has some chemical modifications of a type that is not used in human and bacterial cells? (we don’t want to harm our microbiome) Then those may be usable as a target. 

Another idea is to see what software is needed to help experts in their search for treatment.  Eterna was started in order find better inverse RNA folding algorithms.  It was spectacularly successful with nando’s NEMO algorithm.  Having the right software, for example to try out ‘what-if’ ideas, can make progress much faster.  So what would be needed to help to deal with COVID19?

Along the same lines, if there are several different energy models under consideration, wouldn’t it be possible to have groups of players for each one, trying to replicate the success of Eterna in dealing with the Vienna 1 energy model.  Of course this would require a lot of players, but the hope is that the numbers will come over the next few weeks.

Corona from a math perspective

Exponential growth and epidemics by 3 Blue 1 Brown

However there is something you can do about corona, besides staying at home. :wink:

Folding@home runs protein simulations in relation to different diseases. For this they have asked for computer power from private citizens. Here is a fine introduction. 

At the moment the world has run them out of Covid-19 tasks, but there are still plenty of other diseases needing some calculations. 

Indeed, a coronavirus is a sphere. The unique spike proteins on the surface of SARS-CoV-2 are what make it so effective at attacking its host. The University of Texas rendered the molecular structure of the spike protein. The mRNA vaccine created by Moderna targets the spike protein. The vaccine is currently being tested in volunteers in Seattle to see if it is safe. mRNA therapeutics have a history of problematic side effects.

One possible angle of attack for RNA-based therapy would perhaps be the RISC pathway.


Once the viral genome is delivered inside the cell, it acts like a messenger RNA (mRNA). So it should be possible to down-regulate its expression the same way that normal genes are. There exists methods of siRNA delivery like the one pictured above (the top-left thing is called a “bola”)

siRNAs are structurally very simple. They are dsRNA duplexes of about 20 nts length. Not a very challenging puzzle, but there are certain rules, in particular, the thermodynamics of separating the strands, which determines which of the strands becomes the sense one (normally discarded) and anti-sense (the complementary to the target sequence in the target mRNA). Also, both strands should be BLASTed against the human genome, so as to avoid or mitigate nasty side-effects (off-target matches impacting other genes)

If I recall correctly, siRNA design softwares do exist. I have no idea how good they are, though. Could Eterna do better?

Just 2 cents from an Eterna-retiree :slight_smile:

1 Like

Thank you nando. So good to hear from you. Stay well, stay safe.

Beautiful idea. Would anyone in the community be interested in desigining a test puzzle based on Nando’s idea for Eterna ASAP?

Everyone – we are at a critical time where we are getting ready to deploy the lab online and make experimental preparations in our academic research groups and with industry partners. To start, we will be looking to design mRNAs for some ‘classic’ test proteins like GFP and/or luciferase but we also want to include at least one antigen, even in the first round, that could immediate convert into a vaccine. Help us pick this molecule.  I’ve started a thread over here to discuss:

1 Like

This article suggests targeting the protease.
And the paper published yesterday to which it refers:

Do you have a timeframe for when the first lab will start? Will eteRNA support mobile applications soon?

If I got the labels right…
Here they are in SwissTargetPrediction. Click the links.




What do structures of viruses normally look like?

I have taken a look at the structure of the viruses that David Goodsell has been drawing images off. I sent the images to Rhiju. 

Wow, the Zika is much more messy than coronavirus. But still with some regularly spaced protein molecules on the viral RNA. 

Zika Virus by David Goodsell

Not all of the RNA is wrapped in nucleoproteins. 

In Ebola, I can see some arms from the membrane holding the proteins in the tube at very regular intervals. It is so beautiful.

Ebola Virus, 2014

Ebola Virus Proteins

Ebola looks like it is ordered like a tobacco mosaic virus: Proteins regularly wrapped around the RNA thread. In a spiral.

Tobacco Mosaic virus

Measles has chosen a weird inbetween way of a tubed protein coat holding in a RNA spiral, crossed with a 3D sphere casing:

Measles Virus Proteins, 2019

Measles Virus Proteins

Measles looks like it has its RNA on the outside of the proteins.

HIV is weird. It is a capsid inside a capsid.

HIV in Blood Plasma, 1999

Eli: @rhiju, I wonder if there is a database with images of real virus structures. Sort of like the microscope images that David Goodsell has been drawing after. I’m curious to see how different families of viruses look. Also are you aware of any database over nucleoproteins or nucleosomes? Like proteins used for keeping RNA together inside a viral capsid? 

I recall something with RNA and proteins in PDB also have their chemical modifications listed, so I may be able to look up such modifications on nucleoproteins myself. I am just wondering if there already is some kind of overview.

Rhiju: @eli I don’t know of a great overview of RNPs or virions. Your collection of Goodsell images is the best I’ve seen, actually. I wonder if they link into more detailed docs/reviews with, say, actual micrographs of the virions.

I have added links for the images. Most of the descriptions linked below the images contain references to the papers that the images are based on. 

More thoughts on viral “histones” and epigenetics

Nucleoproteins that bind to RNA or DNA are called respectively ribonucleoproteins and deoxyribonucleoprotein. 

I have found out that epitranscriptome is kind of a coverword for biochemical modifications of RNA itself. 

In the wiki article on the subject one such enzyme editor of RNA chemical modifications was mentioned. ADAR. I then checked for ADAR in relation to viral RNA. ADAR can change transcription rate in HIV. By modifying the viral RNA.

Editing of HIV-1 RNA by the double-stranded RNA deaminase ADAR1 stimulates viral infection

I find this quote from another paper, extremely interesting:

“Influenza A virus (IAV) genomes are composed of eight single-stranded RNA segments that are coated by viral nucleoprotein (NP) molecules. Classically, the interaction between NP and viral RNA (vRNA) is depicted as a uniform pattern of ‘beads on a string’.”

It gets even better with the continuation: 

“Using high-throughput sequencing of RNA isolated by crosslinking immunoprecipitation (HITS-CLIP), we identified the vRNA binding profiles of NP for two H1N1 IAV strains in virions. Contrary to the prevailing model for vRNA packaging, NP does not bind vRNA uniformly in the A/WSN/1933 and A/California/07/2009 strains, but instead each vRNA segment exhibits a unique binding profile, containing sites that are enriched or poor in NP association. Intriguingly, both H1N1 strains have similar yet distinct NP binding profiles despite extensive sequence conservation.”

Genome-wide analysis of influenza viral RNA and nucleoprotein association

Histones can move around a bit, due to how the genome has to get read.

If ribonucleoproteins really behave as histones, they shouldn’t always be in the same spot.

Giant viruses and histones

I wonder how the giant viruses pack their massive genome? They are DNA but some of them have a genome larger than the smallest bacteria.

eli: Ha, I found them. Histones in a Giant virus.

Giant viruses may play an intriguing role in evolution of life on Earth

Those folks even think that eukaryotes are related to giant viruses

“In the study, Erives analyzed the genome of a virus family called Marseilleviridae and found it shares a similar set of genes, called core histones, with eukaryotes.”

Plus I love this description: “Core histones are packagers, like professional gift-wrappers”.

“As he analyzed Marseilleviridae’s genomes in data provided by the National Institutes of Health, Erives noticed the giant virus family encodes the eukaryotic core histones H2B-H2A and H3-H4. Unlike eukaryotes, however, these Marseilleviridae core histones are primitively fused as dimer proteins.”

I’m aware that Giant viruses are DNA viruses. Still I find it really interesting. There are more of these giants that have histones. Medusa has a full set. 

Now viruses have a habit of co-opting a lot of genes. But from what I get, the histones should be conserved and not just temporarily grabbed.

Rhiju: eli, this is really cool.

Vaccine idea

When working on puzzles with the covid-19 coding regions of the proteins, I noticed that it is difficult to increase the number of pairs. This would suggest that the mRNAs coding for the virus proteins are very stable. However, it is easy to decrease the number of pairs, which should result in the mRNA breaking down faster.

Could Eterna design the mRNA coding regions so that more bases are unpaired and therefore more readily exposed to degradation? To be clear, the amino acid sequence and proteins would be unchanged, but the RNA coding regions would be altered. The 10 coding regions could then be packaged together into the full genome creating a replica of the virus, but with properties that allow it to be broken down quickly. That way the body’s immune system can respond without being overwhelmed.


The motivation for trying to make the RNA more stable is for world-wide deployment, especially in underdeveloped countries. As it stands now, the state of the art for RNA vaccines is that they have to be kept refrigerated at all times, from manufacturing to administration, and they have to be put into a syringe just before being injected. The ideal for distribution is that they would be put into a disposable syringe at the factory, and not have to be so carefully refrigerated.

(I may not have these details exactly right; this is my understanding from Rhiju.)