#foldit

[wtf-fun-factss]

After 13 years of trying to find the structure of a protein, scientists created an online competitive game called Foldit that asked gamers to try and solve it. The structure was then found in 3 weeks. – WTF Fun Facts Source: https://www.theguardian.com//technology/2014/jan/25/online-gamers-solving-sciences-biggest-problems

[neurosciencenews]

Citizen Scientists Are Helping Design New Drugs to Combat COVID-19 Thanks to Foldit Game

Feeling at a loose end due to social distancing? You can help researchers design drugs that could help stop coronavirus from infecting human cells. Using a free computer game called Foldit, citizen scientists are helping researchers to hunt for proteins that might stop COVID-19.

[sadlyglyphtaken]

You, too, can design a wrench to throw in the workings of the coronavirus

I pretty much never post about my job here. I do love my field, but it’s not a special interest of mine. Just a normal interest. And while the things I do are pretty easy to explain, it’s hard to explain why they matter.

Now it’s easy to explain why they matter, too. Unfortunately.

The term I’m most comfortable with for my job is protein engineer, but nobody uses that. So my bio says computational protein design specialist instead. I sit at a computer and design proteins. This is important because anything important that has ever happened has been accomplished with the help of proteins. They are the most useful thing to make living things out of, because you can use them as both structural components (struts, ropes, cages, armor plating, nets, pipes, spools) and as machines (assemblers, disassemblers, sorters, recyclers, dynamos, rotary motors, ratchet motors, pumps, locks and keys). They do all the things that make things be alive. (By the way, if you’re on desktop, every one of those words in the parentheses is a link, if you’re curious.)

More precisely, I wish I sat at my computer and designed proteins. Today, instead of doing that, I attended the first half of an emergency protein design conference called to discuss how we can design a protein that can be used to cure the coronavirus. It was five hours long and I’m really tired and I got nothing done today.

Even on a normal day, though, I don’t really design proteins. I just read papers, write programs that make charts, and write apologetic letters to my advisor.

This is because actually designing a protein is the easiest step. The rest is paperwork and tears. Designing a protein is so easy, actually, that I get computers to do it for me. But designing proteins is like playing Starcraft. You can beat a computer just by knowing what the hell you’re doing.

About a decade ago my research community made a program, called Foldit, that let anyone design proteins. We do the paperwork and tears, while you do the design. It’s a fair trade, because design is fun, and we still get paid. (Most of us, anyway.)

Right now, the world only has to accomplish one protein design task. This is the simplest it’s ever been: make a protein that sticks to some part of the spikes on the coronavirus. That’s it. Make the right protein, and you’ve cured the coronavirus.

Everybody in my research community is scrambling head over heels to attack the problem from all angles. With Foldit, you have… somewhere between 10 and 50 percent of those angles available to you. You make something even vaguely successful, and we can use it as fodder for our models and algorithms and clueless neural nets until we’ve aggregated the design for the stickiest motherfucking protein the coronavirus has ever seen. And it won’t be able to do anything to stop us.

Do you want to do something about the coronavirus, and think you can handle protein design? Read below the cut, I’ll explain exactly how you can do that. Do you want to do something about the coronavirus, and don’t think you can handle protein design? Read below the cut anyway, and you’ll change your mind, because protein design is about as hard as sculpting, except with a way lower skill ceiling.

Please help us. We believe in you. We discuss Foldit models at our conferences. We take you seriously. We take this pandemic seriously. We need you.

On board? Let’s dive in.

A protein is a sticky string. There’s two kinds of sticky: orange sticking to orange (”nonpolar” sticky), and red sticking to blue (”polar” sticky). Here’s a short protein, technically so short that we should refer to it as a “peptide”:

You can see that it’s organized into groups of 3, called “residues”: little red bit (a hydrogen), an optional orange bit of some kind (the “side chain”), and a blue bit (a carboxyl). At the ends, the red and blue bits get bigger. The fact that the red and blue bits are always there means that they can be arranged into some simple patterns, assuming there’s room. But let’s talk about the side chains for a moment.

Some side chains end in red and blue bits. These are the “polar” and “charged” side chains, and when the protein tangles itself into the right shape, they’re found on the outside. This is because outside the protein is water, and water molecules have a red part and a blue part, so the side chains want to stick to it. I put a water molecule, to scale, next to the first charged side chain, so you can have a visual. Polar and charged side chains are called “hydrophilic” side chains for this reason.

The pure orange “nonpolar” side chains, meanwhile, gather on the inside, sticking to each other in a gooey mess. They’re the opposite of hydrophilic: hydrophobic. When you achieve that gooey mess, you have “buried your hydrophobics”.

90% of figuring out what shape a protein will tangle up (euphemistically “fold”) into is burying your hydrophobics.

There’s only 20 kinds of side chains in total, and that’s also counting the weird residues with not-quite side chains like proline, that weird bend in the last but one residue in our little peptide, and glycine, the residue with no side chain at all. (Besides a measly hydrogen, and everyone knows hydrogens aren’t real. That’s a computational protein engineering injoke; feel free to use it. The joke is nobody knows exactly where hydrogens are, and nobody cares.) Here’s a family portrait:

I’m not gonna explain any of the weird ones here, because honestly, it’s not super important. Red sticks to blue, orange sticks to orange, and stuff that looks like it’s connected is connected. That’s it, that’s proteins. I guess it’s worth mentioning that the yellow parts are sulfur, but sulfurs are just large, bratty oxygens (the blue bits). They’re not special.

Remember how I said the regular alternation of red bits and blue bits on residues creates some simple, reliable patterns? “Some” is two. Well, technically a few more than two, but nobody really pays attention to the other ones, because they’re just knockoffs of the first two. Here is the first pattern. It’s the favorite pattern of isolated bits of protein with nothing better to do:

This is called an alpha helix. They’re stiff, like rods. You’ll see a lot of these. The other one happens when you lay bits of proteins side by side:

This is called a beta sheet. If you leave a bucket of any protein whatsoever in one place long enough, eventually it’ll fall apart and turn into these. That’s part of why cells have to clean their proteins up: otherwise, it’ll turn into this junk. (”This junk”, on the cellular level, is called amyloid fiber, and is deadly to the cell. On the human scale, it’s called silk, and it’s nice and soft.) But this is also a very structurally useful pattern; it creates tight but flexible weaves. They get shaped into barrels or parts of barrels, usually, but sometimes they get used as walls instead. And sometimes they’re very small and consist of just two short bits that need to get glued together at the sides.

The rest of the protein is just spaghetti, euphemistically called “loops”. The functional bits that attach to stuff outside the protein tend to be on the loops, because the loops are the most flexible and can be posed the most precisely. In real life, proteins are constantly undergoing vibrating motions on every level. The biggest motions are also the slowest; pac-man-shaped proteins will “breathe”, opening and closing at regular intervals. Usually the only time they’ll be able to bind stuff is when they’re open. The takeaway here is that all the stuff inside the protein is gonna wiggle anyway, and you have to make sure it looks the best when it’s wiggled open, if you’re trying to stick something.

Congratulations. That’s everything I know about how proteins get their shapes, pretty much. Your mission is now to make a protein that sticks really well to some important part of the coronavirus’s spikes. The spikes are made of protein, and when they get near an alveolar cell, they open up and jam a lockpick into a “lock” protein called ACE2 on the surface of the cell, which makes the cell sluuurrrp up the virus. Some good ideas are to cover the lockpick so it can’t activate the lock, or to jam the opening and closing mechanism so the spike’s shape gets distorted.

There’s also a couple of other proteins the coronavirus makes that you could try to mess up. A popular target is its two proteases. Viral proteins come like the plastic bits for assemblable toy models: all the pieces are there, but they’re connected together, and you have to break them apart before you can put them together. That’s what the proteases do. (The second protease also cuts off the “to be destroyed” tags that the cell’s proteins keep putting on the viral proteins.) It’s much harder to get your protein drug inside the cells, but once they’re there, they can be as effective as a drug that targets the virus before it gets into the cell. Leave the drug delivery to us. (Or to big pharma, idk. I got a C+ in drug delivery in undergrad, and got it bumped up to a B when I reminded the professor that my attendance grade was low because I was in the psych ward for a chunk of the semester. That’s still more drug delivery than most protein engineers have taken.)

You’re ready!

Here’s a “Getting started with Foldit” tutorial.

Here’s an in-depth analysis of the coronavirus spike for Foldit players, with an accounting of some of the more advanced tools at your disposal. (I highly recommend DDG, although it’s slow to compute in real time.)

[viscardiac]

started playing foldit and i'm suffering but i shall SUCCEED

[kimjameson]

Folding is my superpower and, no, I’ll never part with that little red wagon/tp holder. #foldit #crushit #makethemostofit #littleredwagon #bambootp #chenillebedspread (at Columbia, Tennessee) https://www.instagram.com/p/CPVlqethZyT/?utm_medium=tumblr

[tessellated-lightning]

Ever fancied yourself a research scientist on a quest to discover a cure for a serious disease ? Don’t want to spend 10 years writing papers about the reproductive cycle of naked mole rats ? Now you can !

Yes, for real. Thanks to the magic of Foldit !

Foldit is a protein folding game whose scoring system is built on top of the Rosetta@home * distributed computing** project for protein structure prediction.

It then takes the highest score and submit them for further analysis by real research institutions as public domain knowledge.

Another aim of the game is to see if human intuition can lead to better results than those achieved by computer alone. The best strategy employed by human could then be taught to computers.

It is still active, there was a recent wave of publicity about it due to the Coronavirus:

Like here: https://www.freethink.com/articles/coronavirus-antiviral-medications

And here: https://www.geekwire.com/2020/protein-puzzle-game-called-foldit-turns-99-promising-ways-confound-coronavirus/

And here: https://www.the-scientist.com/news-opinion/scientists-use-online-game-to-research-covid-19-treatment-67230

Tip: The tutorial levels are supposed to be relatively easy, if you find yourself having a really hard time, you are probably over-thinking it. These levels are there to illustrate the tools of the game. The basic pattern is generally pretty high-scoring potential, and in my experience, trying to get creative during the tutorials without in-depth in-game expertise often results in more frustration than it is worth.

___________________________________________________

Inventors discovery rights: 

“All significant scientific discoveries (such as structures, algorithms, etc) made in-game will be made publicly available. Discoveries will be governed by US patent law and handled by the University of Washington Center for Commercialization. Individual players who contributed to the discovery will be considered inventors for the purposes of discovery ownership.”

___________________________________________________

distributed computing project for protein structure prediction:

What this means is that the scoring system is based of a program running calculations to find the most efficient ways to fold proteins for research progress, based on criteria such as energy-efficiency and likelihood to appear in nature. (aka “proven and true” structures)

Now the cool thing about this program is that it relies on distributed computing: What this means is that instead of merely using it’s own power to run the calculations, it use the spare computing power of any users who wish to join the project. The potential that would have been otherwise wasted.** The Rosetta program requires no efforts beyond installation/setup, and is another good way to contribute to research.

___________________________________________________

* I first heard of the Rosetta project in a scientific review when I was... 14 ?, but sadly my computer was extremely slow and I couldn’t contribute. I had completely forgotten the name and thought I missed the occasion by a decade. 

I guess it’s never too late.

** A bit like the weeping angels in Doctor Who, who sends you into another time to harvest the potential energy of your unlived lifetime, except minus the part where you are forced to, and the part where you do not get to see the people you love ever again for the rest of your life.

- “Fascinating race, the Weeping Angels. The only psychopaths in the universe to kill you nicely.” 

[myesportsglobe]

Fight #coronoa with foldit! Read more here to find out how gamers can help fight Coronavirus! https://myesportsglobe.com/games/foldit-create-covid-19-vaccine #foldit #fight #coronavirüs #covid #covid19 #fightingcovid #greatgame #allgamers #gamersunited #gamerlife #life #dna #virus #goodnews #togetherathome #people #myesportsglobe #esports #gaming #latestnews #news #bhagcorona #standunited #stayhome https://www.instagram.com/p/B-_gDUFAWcO/?igshid=ehkyyu6yjal3

[shuga-hill]

This puzzle game lets you help create a coronavirus vaccine

Foldit allows gamers to take a shot at helping create a coronavirus vaccine via a game of protein folding -- and it's free-to-play.