74 replies [Last post]
agcohn821's picture
User offline. Last seen 3 hours 51 min ago. Offline
Joined: 11/05/2019
Groups: Foldit Staff
Newsletter June 25: We've Been Busy!

Hey folders!

Dev Josh here with your weekly Foldit update.

Solutions from This Week's Puzzles

(Disclaimer: This is not scientific feedback; these solutions are not officially endorsed by the Foldit scientists.)

Puzzle 2005: IL-2R Binder Design

Puzzle 2006: Symmetric D2 Tetramer Design: Limited Interface

Puzzle 2007: Large Electron Density Transcription factor

Upcoming Updates

Here’s what the devs have been up to lately:

  • New interns! Our team has two new interns from RosettaCommons, Foldit’s larger software development community. (Rosetta is the software Foldit is built on!) Our interns will be working on special new levels and contests, major server improvements to Dojo mode, and a bunch of new levels, tweaks, and updates to Dojo mode.
  • The great mergening! We’re doing some big changes on our backend to update Foldit with some of the latest and greatest tools from Rosetta. Not only does this mean our development process will get a much-needed tune-up, we’ll also have some new tools available to work with. If you’ve been following the news on trRosetta, or if you like the idea of adding deep neural networks to your folding, this is the part where you squeal with delight.
  • Better help. We’re polishing the in-game help panel (which has seen improvements since this wiki page was updated) to be even more flexible and helpful.
  • Bringing back the R. We’re really sorry that the R hotkey for removing bands was taken over by the Remote Control feature. We’re bringing back your hotkeys, for real this time.
  • A whole new perspective! Soon you’ll be able to check your fold in an orthographic view! Thanks jeff101 for the suggestion!

Wow, that’s a ton! Bear with us while we try to roll out all of these cool new improvements. It might take us a while to get them all tested and ready to go.

Today’s Master Folding Tips

Do you have trouble with getting the hydrogen bonds right on your sheets? Here’s some diagrams and explanations on the geometry of sheet bonding.

There’s a lot you can learn from your fellow players. Join a group and share your solutions! Both the sharer and the evolver tend to learn something from the process.

Until next time, happy folding!

agcohn821's picture
User offline. Last seen 3 hours 51 min ago. Offline
Joined: 11/05/2019
Groups: Foldit Staff
Newsletter July 2: Another Year, Another Newsletter

Hey folders!

Dev Josh here with your weekly Foldit update. Can you believe I’ve been writing these newsletters for an entire year now? The first newsletter was July 3, 2020, before binding metrics were even added to the game! What a long, strange trip it’s been.

Solutions from This Week's Puzzles

(Disclaimer: This is not scientific feedback; these solutions are not officially endorsed by the Foldit scientists.)

Puzzle 2008: TGF Receptor Binder Design

Puzzle 2009: Symmetric D2 Tetramer Design: H-bond Networks

Puzzle 2010: Bos Taurus

Recipe of the Week

For the veteran folders out there, LociOiling has added a bugfix to Bruno Kestemont’s Banded Worm Pairs (BWP) in versions 3.4.8 and 1.4.9.

Player of the Week

A huge congratulations and thanks to NinjaGreg! Not only has his design of the month been synthesized in the wet lab, but he has also been volunteering with the Foldit dev team to help us fix bugs and add new features! You rock, Greg.

Today’s Master Folding Tips

Given the new updates to Banded Worm Pairs (BWP), let’s break down what exactly this recipe is doing and how to use it. This is a very late game recipe for squeezing out just a couple of extra points at the very end of a puzzle. The original BWP was actually written by KarenCH and based on an even older “worm” wiggle by rav3n_pl. Since then, a number of worm recipes have spawned based on this approach. They’re crawling all over the place!

BWP uses bands and local wiggling to try to optimize your fold. The recipe usage itself is pretty simple. The slider adjusts how often you want to save progress, ranging from every time you gain 0.3 points to 1000 points from the recipe. And the textbox is for a name for the save file. These saves are stored in the Menu under Open/Share Solutions. In the later versions of this recipe, it also accounts for objectives like H-Bond Networks, and will run infinitely until you tell it to stop.

Want us to keep talking about recipes, or something else? Post your feedback on our feedback forum!

Until next time, happy folding!

agcohn821's picture
User offline. Last seen 3 hours 51 min ago. Offline
Joined: 11/05/2019
Groups: Foldit Staff
Newsletter July 9: It's Atomic!

Hey folders!

Dev Josh here with your weekly Foldit update.

Solutions from This Week's Puzzles

(Disclaimer: This is not scientific feedback; these solutions are not officially endorsed by the Foldit scientists.)

Puzzle 2011: Cortisol Ligand Binder Design

Puzzle 2012: Symmetric D2 Tetramer Design: H-bond Networks

Puzzle 2013: Scorpion Toxin

Player of the Week

Did you know that every week, LociOiling collects even more screenshots from the top puzzle results and posts them to the Foldit Wiki? What a champ, thanks Loci! Be sure to check out those results so you can see what the top folders are doing in every puzzle!

Today’s Master Folding Tips

Even though Foldit colors entire sidechains as hydrophobic or hydrophilic, this is actually a property of the individual atoms that make up a sidechain! Carbons are hydrophobic, while polar atoms like oxygen and nitrogen are hydrophilic. Keep this in mind when packing your interfaces and cores: it’s not just what sidechains you use, but how you angle them! Hydrophobicity is atomic!

We’re making some major improvements to Dojo mode! Be sure to check it out and let us know what you think on the forums or the #dojo channel of our Discord server!

Until next time, happy folding!

agcohn821's picture
User offline. Last seen 3 hours 51 min ago. Offline
Joined: 11/05/2019
Groups: Foldit Staff
Newsletter July 16: The Art of a Good Wiggle

Hey folders!

Dev Josh here with your weekly Foldit update.

Solutions from This Week's Puzzles

(Disclaimer: This is not scientific feedback; these solutions are not officially endorsed by the Foldit scientists.)

Puzzle 2014: TGF Receptor Binder Design

Puzzle 2015: Symmetric D2 Tetramer Design: H-bond Networks

Puzzle 2016: Nucleosome Protein

Recipe of the Week

Wiggle Shake Wiggle by SaveVMK and Pikamander2

Art of the Week

This week’s art comes from Nicm25! Who knew bands could be such a great artistic tool?

Today’s Master Folding Tips

Moving stuff around in Foldit can be hard. That’s why I made this guide to learn about controlling your movements in 3D space!

Do you have a recipe or screenshot to share? Send it in to mail.fold.it@gmail.com!

Until next time, happy folding!

agcohn821's picture
User offline. Last seen 3 hours 51 min ago. Offline
Joined: 11/05/2019
Groups: Foldit Staff
Newsletter 7/23/21: So Exciting, I Can’t Talk About It

Hey folders!

Dev Josh here with your weekly Foldit update.

Solutions from This Week's Puzzles

(Disclaimer: This is not scientific feedback; these solutions are not officially endorsed by the Foldit scientists.)

Puzzle 2017: IL-2R Binder Design

Puzzle 2018: Symmetric D2 Tetramer Design: H-bond Networks

Puzzle 2019: Crystallin

Upcoming Updates

We’ve got some seriously exciting new things happening just around the corner! But I will let bkoep talk about it more in the next lab report. Instead, I’ll simply share some cool new scientific papers we’ve all got our eyes on:

The Baker Lab at UW has developed a protein prediction algorithm called RoseTTAFold in their latest paper Accurate prediction of protein structures and interactions using a three-track neural network. Meanwhile, literally published on the same day, DeepMind finally describes their own algorithm, AlphaFold, in their article Highly accurate protein structure prediction with AlphaFold. Could this mean we have brand new technologies for quickly identifying whether a fold will be successful? Only time will tell!

Art of the Week

The Olympic, by Nicm25

Today’s Master Folding Tips

Do you tend to cut your protein apart at the loops, but then you can’t get your loops ideal when you merge them back together? Try cutting in the middle of your sheets and helices instead! This preserves your loops better and makes it easier to idealize loops after merging.

Make sure you subscribe to our YouTube channel and click that bell icon so you get notified when our lab reports go live! You won’t want to miss this!

Until next time, happy folding!

agcohn821's picture
User offline. Last seen 3 hours 51 min ago. Offline
Joined: 11/05/2019
Groups: Foldit Staff
Newsletter July 30: Tools for Schools

Hey folders!

Dev Josh here with your weekly Foldit update.

Solutions from This Week's Puzzles

(Disclaimer: This is not scientific feedback; these solutions are not officially endorsed by the Foldit scientists.)

Puzzle 2020: Olivetolic Acid Ligand Binder Design

Puzzle 2021: Symmetric Trimer Design: H-bond Networks

Puzzle 2022: Citric Acid Cycle

Recipe of the Week

Is your cookbook a little sparse? Check out this Toolbox by MurloW!

Tips for School

Educators: If you’re thinking about putting Foldit in your classroom, check out our Educational mode! If you do use Foldit in the classroom, we recommend using it as optional extra credit or a group lab segment. And if you want to give a grade, we recommend grading based on the number of levels beaten or the highest score earned in a science puzzle, not the highest score earned in a tutorial puzzle.

Students: Have a question about Foldit or protein science? Join our Discord where you can ask all the questions you have! And be sure to check the wiki for guides and walkthroughs!

Today’s Master Folding Tips

You can use the Auto-SS tool to figure out if your assigned secondary structures are likely to fold up that way! The secondary structures you assign --- helices, sheets, and loops --- are only suggestions for Foldit’s tools. What really determines their shape is the backbone angles. Auto-SS will tell you, based on your fold, what secondary structures your design will actually take!

Want to see some really great folding? Check out July’s Design of the Month!

Until next time, happy folding!

agcohn821's picture
User offline. Last seen 3 hours 51 min ago. Offline
Joined: 11/05/2019
Groups: Foldit Staff
Newsletter August 6: AlphaFoldit!

Hey folders!

Dev Josh here with your weekly Foldit update.

Solutions from This Week's Puzzles

(Disclaimer: This is not scientific feedback; these solutions are not officially endorsed by the Foldit scientists.)

Puzzle 2023: TGF Receptor Binder Design

Puzzle 2024: Symmetric Trimer Design: H-bond Networks

Puzzle 2025: Polycystein

Recipe of the Week

This week’s recipe is a modification of Fracture by MurloW to include Remixing! Thanks to Artoria2e5 for this remixed Fracture! Fracture is an excellent multi-use recipe for everything from early to late game. It’s on the high-end of average complexity because there are many options if you’d like them, but the primary dialog is easy enough to use. Make sure to open up the recipe output for a quick guide to its usage: essentially, you check whether you’re doing early, mid, or late game, and which preset script you want. The options include DRW, Acid Tweeker, RR (Rainbow Rebuilder), and now some remixing!

Upcoming Updates

  • Enjoying the new AlphaFold tool? Soon you’ll be able to submit up to five solutions at a time!
  • Education mode is getting integrated into Foldit! This Fall, you’ll be able to learn in the Education mode and solve puzzles for science in the same game!

Today’s Master Folding Tips

Curious about the new AlphaFold tool? I know I am! You can read all about its workings in bkoep’s blog post, but here’s the gist: in certain puzzles, you can upload your solutions to AlphaFold -- right next to the button you’d use to upload solutions to scientists or group members. After our servers process your solution, you’ll get three things:

  1. AlphaFold’s solution - given your sequence of amino acids, this is how AlphaFold thinks those amino acids will fold up in the lab. Note that to really see what the solution looks like, you’ll have to “relax” it: that means giving it a shake and wiggle, maybe even with increasing clashing importance like you learned in the Control Over Clashing tutorial puzzle.
  2. Confidence - this is how confident AlphaFold is in its own solution. But because AlphaFold confidence correlates with success in the real lab, this is a good indicator of how likely this sequence will actually fold up well! That’s why we’re asking for players to find solutions with at least 80% confidence. Why does confidence correlate with wet lab success? It’s because a high confidence means there aren’t a lot of decoy folds, or other ways the protein could fold up. Check out this blog post for more about the energy landscape of protein design and why decoys create problems for designing proteins!
  3. Similarity - this is a measure of how similar your solution is to AlphaFold’s solution. If your similarity is low, that means AlphaFold found a totally different way to fold the same amino acids as you! This doesn’t mean much for the success of the protein, but it sure does feel good when AlphaFold agrees with you!

So how do you use AlphaFold well? Notice that AlphaFold’s solution and confidence are based entirely on the sequence of amino acids alone -- unless you mutate, insert, or delete a segment, you’ll always get the same result back. If your confidence is less than 80%, you’ll need to mutate, insert, or delete at least something to make a change. Expert player Susume offers a few tips here:

  1. Make sure every edge strand (the edge of your sheets) has at least one, preferably two, hydrophobic (orange) residues in the core.
  2. If there’s a helix with 5+ hydrophilic (blue) residues in a row, add some oranges in there to become part of the core.
  3. Make sure your loops are ideal and made of residues that are common for that loop. If you’re really ready to dig into the biochemistry of things, check this supplementary material on loop designs from Lin et al. 2015.

Ready to try out AlphaFold? Play puzzle 2027 now and build something great! Our scientists are looking for anything with a high confidence score, no matter how crazy! Try to break AlphaFold, or try to design something spectacular!

Until next time, happy folding!

agcohn821's picture
User offline. Last seen 3 hours 51 min ago. Offline
Joined: 11/05/2019
Groups: Foldit Staff
Newsletter August 13: AlphaFold Tips from an Award-Winning Newsl

Hey folders!

Dev Josh here with your weekly Foldit update.

Solutions from This Week's Puzzles

(Disclaimer: This is not scientific feedback; these solutions are not officially endorsed by the Foldit scientists.)

Puzzle 2026: IL-2R Binder Design

Developer News

This week, our devs were (virtually) attending RosettaCon, an annual conference for Rosetta, our underlying biochemical software. They talked about some recent scientific progress in protein prediction using neural networks, combining machine learning with Rosetta’s modeling, and more! On a more personal note, my summer intern mentee Alex won an award for his poster and I won an award for these newsletters! I look forward to bringing you more award-winning newsletters in the weeks to come :)

ColabFold

For folks interested in a deeper overview of RoseTTAFold and AlphaFold, check out this talk on ColabFold, which makes these tools accessible through Google Colab.

Today’s Master Folding Tips

Veteran player Susume offers some insights on how to interpret your AlphaFold results: low confidence from AF means that your sequence (of amino acids) doesn’t strongly select a single fold -- it could easily misfold or fail to fold. So, in order to improve AF confidence, you need to either make the sequence more realistic or more selective for a single fold.

Today, I’m going to focus on the second one: what happens if your fold isn’t selective? We’ve talked before about how decoys can lead to misfolding. The AlphaFold result you get back might very well be one of those decoys! What can you change about your fold to make that decoy less desirable? For example, this could mean:

  • Hydrogen bonds in your fold that are BUNS in the decoy
  • Hydrophobic oranges in the core of your fold that are exposed in the decoy
  • Sidechains that pack nicely in your fold, but are clashes or voids in the decoy

Try out the AlphaFold puzzle now!

Until next time, happy folding!

agcohn821's picture
User offline. Last seen 3 hours 51 min ago. Offline
Joined: 11/05/2019
Groups: Foldit Staff
Newsletter August 20: AlphaFold Tips with an Eterna Analogy

Hey folders!

Dev Josh here with your weekly Foldit update.

Solutions from This Week's Puzzles

(Disclaimer: This is not scientific feedback; these solutions are not officially endorsed by the Foldit scientists.)

Puzzle 2028: Platypus Venom

Puzzle 2029: TGF Receptor Binder Design with AlphaFold Predictions

Puzzle 2030: Antifreeze Protein

Today’s Master Folding Tips

Today’s tips are a continuation of last week’s discussion on avoiding decoys. If you’ve played Eterna, you might know that they have a Target mode and a Natural mode. The Natural mode is also called the minimum free energy structure, or the secondary structure with the lowest free energy. This is exactly like the secondary structure (SS) in Foldit with the highest score. The Target, then, is the desired SS. Another way to think of this is that Target is the structure you want to form, and Natural is the structure that is most likely to form. Your goal in Eterna is to get the Natural mode to match the Target, which you can either do by lowering the free energy of the Target (raising the score of the Target fold), or raising the free energy of the Natural structure (lowering the score of the Natural fold).

Why am I talking about this? Because when you fold with AlphaFold predictions, AlphaFold is showing you what it predicts to be the Natural solution, to use Eterna’s terminology. It’s showing you the most likely fold. And your design is your own Target. So what do you do if your proposed structure doesn’t match the likely outcome? You have two choices: make your structure better, or make the likely outcome worse. For example, if AlphaFold predicts that it will misfold because it found a hydrogen bond you didn’t anticipate, take that hydrogen bond out! Mutate one of the hydrophilics into something that won’t form that bond in AlphaFold’s structure, and will only make hydrogen bonds in your designed structure.

Try it out in the latest Symmetric Trimer Design with AlphaFold Predictions! And if you find any more strategies, let me know so I can share them with the rest of you!

Until next time, happy folding!

agcohn821's picture
User offline. Last seen 3 hours 51 min ago. Offline
Joined: 11/05/2019
Groups: Foldit Staff
Newsletter August 27: Getting Edgy with Sheets

Hey folders!

Dev Josh here with your weekly Foldit update.

Solutions from This Week's Puzzles

(Disclaimer: This is not scientific feedback; these solutions are not officially endorsed by the Foldit scientists.)

Puzzle 2031: IL-2R Binder Design with AlphaFold Predictions

Puzzle 2032: Symmetric Trimer Design with AlphaFold Predictions

Today’s Master Folding Tips

Let’s talk about sheets and strands! Although we usually use “sheets” to refer to any of the zigzag bits, the scientific term for what we usually think of as a sheet is a “strand,” and a “sheet” refers to the group of strands all hydrogen bonded together.

Now that we have that language, I can tell you more about edge strands. Foldit likes it when the strands on the edges of your sheets are blue hydrophilics. But in the real lab, those edges are too floppy without some sticky oranges to pull the edges into the core of the protein.

What can you do about this? Well, besides adding some oranges to your edge strands, you can also make the edge strands shorter. This makes them less likely to be floppy loops. You can also try adding a beta bulge --- an extra residue in the normally alternating zigzag of a strand, which causes the strand to bulge a little. You can read more about beta bulges (and glycine kinks, which I haven’t mentioned) here, or on Wikipedia, or in the advanced section of my guide to Foldit. For the more expert folders, check out figure 3 of this paper! Thanks to veteran player and up-and-coming scientist Susume for all of these tips!

Are you interested in using Foldit for education? Tell us about it at mail.fold.it@gmail.com! We’re actively looking into ways to support education using Foldit!

Until next time, happy folding!

agcohn821's picture
User offline. Last seen 3 hours 51 min ago. Offline
Joined: 11/05/2019
Groups: Foldit Staff
Newsletter September 3: Back to School with Foldit!

Hey folders!

Dev Josh here with your weekly Foldit update.

Solutions from This Week's Puzzles

(Disclaimer: This is not scientific feedback; these solutions are not officially endorsed by the Foldit scientists.)

Puzzle 2034: TGF Receptor Binder Design with AlphaFold

Puzzle 2035: Symmetric Trimer Design with AlphaFold

Upcoming Updates

It’s September and you know what that means: time to head back to school! This back-to-school season, we’re integrating Educational Mode directly into the regular Foldit client! We’ve also made some big improvements to Educational Mode, including two brand new sets of puzzles: the Popular Proteins set lets you play with several of nature’s classics, including collagen, insulin, ubiquitin, and fibrin! The Bacteria and Viruses set challenges you to play with some of our most difficult puzzles over the years as we tackled Coronavirus, Ebola, the HIV protease, E. Coli, and the famous Mason-Pfizer Monkey Virus!

On top of that, here are some other minor features you can look forward to in the upcoming update:

  • New camera features! Try pressing Shift+Home to make the camera rock back and forth gently, or Alt+Home to spin the camera around.
  • Tooltips in the Campaign and Education puzzles now support going back to previous tooltips and quickly closing them out completely.
  • You can now press Tab on a residue in the Campaign and Education puzzles to view additional information on that residue.

Are you an educator using Foldit in your classroom? Let us know, and give us feedback on how we’re doing! Email us at mail.fold.it@gmail.com to tell us how you’re using Foldit and what we can change to better suit your educational needs.

Today’s Master Folding Tips

Today we’re taking a closer look at the amino acids (AAs) in your toolbox: there are 20 of them and they each have unique properties for folding. For now, though, I’m going to simplify it a lot into their most common secondary structures.

For helixes, just remember MALEK: Methionine (M), Alanine (A), Leucine (L), Glutamate (E), and Lysine (K). Those are the AAs that love being in helixes.

For sheets, it’s less clear what to put, but we know that Isoleucine (I), Valine (V), and Glutamine (Q) do pretty well there.

For loops, your best bet is usually Glycine (G), Arginine (R), or Serine (S), although Threonine (T) likes sitting next to Serines if you have one.

Next week, we’ll continue our close look at some more AAs and their properties

Reading this newsletter on the forums? Sign up here to get them delivered straight to your email a few days earlier!

Until next time, happy folding!

agcohn821's picture
User offline. Last seen 3 hours 51 min ago. Offline
Joined: 11/05/2019
Groups: Foldit Staff
Newsletter September 10: Check Out Those Bonds!

Hey folders!

Dev Josh here with your weekly Foldit update.

Solutions from This Week's Puzzles

(Disclaimer: This is not scientific feedback; these solutions are not officially endorsed by the Foldit scientists.)

Puzzle 2037: Cortisol Ligand Binder Design with AlphaFold Predictions

Puzzle 2038: Symmetric Tetramer Design with AlphaFold Predictions

Take a close look at that hydrogen bond network! That’s all arginine and glutamate putting in some serious work! We’ll dive more into them in today’s tips.

Upcoming Updates

Here’s what the dev team has been working on recently!

  • A new-and-improved interface for getting predictions from AlphaFold
  • New features to help with Electron Density puzzles

We also have a few projects in the pipeline for some bigger changes to the interface, but we’ll talk more about that later!

Today’s Master Folding Tips

Today we’re continuing our exploration of the 20 amino acids in your toolkit! Starting with...

Alanine: swap this little guy in anywhere, he’ll fit in any SS!

Arginine: this hydrophilic has a lot of rotamers, or ways it can be rotated, so put it anywhere on the outside of your protein and shake it out!

Asparagine: This sidechain makes good hydrogen bonds with the backbone, so it’s useful at the edges of helices and when your sheets are turning.

Aspartate: Do not put this one in helices!

Cysteine: If you can get two cysteines together, they form a disulfide bridge, which is more powerful than a hydrogen bond! Use a pair of these at right angles when you need a really strong bond.

Glutamine: This long hydrophilic is good at stabilizing helices and sheets, but less good at being flexible. Keep it out of loops.

Glutamate: This one is best for helices and hydrogen bonds.

Glycine: The most flexible amino acid of them all, it’s best for loops (especially U-turns) and terrible at stabilizing helices and sheets. If a residue is having trouble bending, put a glycine next to it to take some of the load off.

Well! I think that’s enough for today. Ready to try your hand at designing with these amino acids? Try Puzzle 2041: Symmetric Tetramer Design with AlphaFold Predictions now!

Until next time, happy folding!

agcohn821's picture
User offline. Last seen 3 hours 51 min ago. Offline
Joined: 11/05/2019
Groups: Foldit Staff
Newsletter September 17: Pumpkin Pi Stacking

Hey folders!

Dev Josh here with your weekly Foldit update. Today’s theme is pi stacking. And because it's fall, it's time for pumpkin pi stacking! Read more about pi stacks here. In the meantime, enjoy this visual pun:

Solutions from This Week's Puzzles

(Disclaimer: This is not scientific feedback; these solutions are not officially endorsed by the Foldit scientists.)

Puzzle 2040: IL-2R Binder Design with AlphaFold Predictions

Puzzle 2041: Symmetric Tetramer Design with AlphaFold Predictions

Today’s Master Folding Tips

This week we’re continuing our deep dive into the amino acids and how to use them. Let’s jump back into it:

Histidine: This hydrophilic has both a donor and an acceptor. That, plus its uniquely pentagonal aromatic ring makes it a versatile sidechain. Despite its hydrophilicity, you can use it for pi stacking! Try stacking it with other histidines or with the hydrophobic aromatics.

Isoleucine: This AA is a fan-favorite for sheets and terrible in helices! Fill your sheets with isoleucine to start with, and mutate in some valines, leucines, and methionines as needed.

Leucine: Despite being similar in shape and function to isoleucine and valine, this branched-chain AA is different enough that it prefers helices over sheets, but keep it in the hydrophobic core!

Lysine: This long AA has two great purposes: (1) forming hydrogen bonds, especially with negatively charged AAs like Aspartate, and (2) stabilizing helices. Keep in mind, though, that its long carbon chain is hydrophobic: in the ideal case, only the charged end is sticking out of the protein.

Methionine: This hydrophobic does wonders for stabilizing a helix, but it still wants to be in the hydrophobic core if it can be. Although it has a sulfur, Methionine is pretty non-reactive and can’t form disulfide bridges like cysteine can.

Phenylalanine: Also known as the queen of pi stacking. If you need your pi stack to hydrogen bond as well, swap in a tyrosine instead. Otherwise, keep the phenylalanine to avoid BUNS.

What an info dump! Take a week to absorb this, and I’ll be back again next Friday with more tips on using your AAs to the best of their abilities!

Until next time, happy folding!

agcohn821's picture
User offline. Last seen 3 hours 51 min ago. Offline
Joined: 11/05/2019
Groups: Foldit Staff
Newsletter September 24: PST, Here's How To Turn

Hey folders!

Dev Josh here with your weekly Foldit update.

Solutions from This Week's Puzzles

(Disclaimer: This is not scientific feedback; these solutions are not officially endorsed by the Foldit scientists.)

Puzzle 2043: CD47 Binder Design

Puzzle 2044: Symmetric Tetramer Design with AlphaFold Predictions

Today’s Master Folding Tips

This week, we’re wrapping up our review of the amino acids in detail!

Proline: Master of bending, this amino acid has a niche purpose but excels at that one thing. Use it to link strands together in a sheet or put it in the first turn of a helix to help the helix get started, but never put it in the middle of helices or sheets.

Serine: Because of its small size, serine is often useful in loops and less useful in helices. Serine is often found next to Threonine forming ST motifs and ST turns.

Threonine: Serine’s buddy for ST motifs and ST turns. Put it in loops. The two of them are also useful for making hydrogen bonds.

Tryptophan: Hoo boy. This guy is large. Literally the largest amino acid of them all, and the only amino acid with two rings. Tryptophan is useful for pi stacking and filling up space in the hydrophobic core.

Tyrosine: Like phenylalanine, tyrosine can be used for pi stacking. But unlike phenylalanine, it has an oxygen hybrid acceptor/donor, making it great when you want to pi stack but also need to satisfy a buried unsatisfied polar atom in that space.

Valine: Like its cousins isoleucine and leucine, Valine is most common in sheets and terrible in helices because it’s C-beta branched and bulky.

Want to see what a great design looks like? Check out September’s Design of the Month by spvincent and ichwilldiesennamen!

Until next time, happy folding!

agcohn821's picture
User offline. Last seen 3 hours 51 min ago. Offline
Joined: 11/05/2019
Groups: Foldit Staff
Newsletter October 1: The No BUNS Challenge

Hey folders!

Dev Josh here with your weekly Foldit update.

Solutions from This Week's Puzzles

(Disclaimer: This is not scientific feedback; these solutions are not officially endorsed by the Foldit scientists.)

Puzzle 2046: TGF Receptor Binder Design

Puzzle 2047: Symmetric Tetramer Design with AlphaFold Predictions

newsletter_2047.png

The No BUNS Challenge

Looking to push your Foldit game to the next level? This month, try our No BUNS Challenge! On any design puzzle, create a solution with absolutely zero* buried unsatisfied polar atoms (BUNS) and share it with scientists with the tag “NO BUNS” to receive a special role on our Discord server! Good luck, BUNS-fighters!

*Some puzzles may have existing BUNS that are impossible to satisfy. Players will win the challenge for submitting a solution with the minimum possible number of BUNS per puzzle.

Today’s Master Folding Tips

Here are some tips for finding and satisfying BUNS:

Turn on “Show bondable atoms” to light up your fold like a jigsaw puzzle of connect-the-dots

Show hydrogens (“Show Bondable H”) for an even clearer picture of potential bonding. You can also see this gallery and this table for more details on which amino acids can form what bonds.

Struggling to get a sidechain in the right rotation to form a bond? Use the Pick sidechains tool to cycle through rotamers.

Want to get these newsletters directly in your email as soon as they go out? Click here to sign-up and get your weekly Foldit news and tips every Friday!

Until next time, happy folding!

agcohn821's picture
User offline. Last seen 3 hours 51 min ago. Offline
Joined: 11/05/2019
Groups: Foldit Staff
Newsletter October 1: The No BUNS Challenge

Hey folders!

Dev Josh here with your weekly Foldit update.

Solutions from This Week's Puzzles

(Disclaimer: This is not scientific feedback; these solutions are not officially endorsed by the Foldit scientists.)

Puzzle 2046: TGF Receptor Binder Design

Puzzle 2047: Symmetric Tetramer Design with AlphaFold Predictions

The No BUNS Challenge

Looking to push your Foldit game to the next level? This month, try our No BUNS Challenge! On any design puzzle, create a solution with absolutely zero* buried unsatisfied polar atoms (BUNS) and share it with scientists with the tag “NO BUNS” to receive a special role on our Discord server! Good luck, BUNS-fighters!

*Some puzzles may have existing BUNS that are impossible to satisfy. Players will win the challenge for submitting a solution with the minimum possible number of BUNS per puzzle.

Today’s Master Folding Tips

Here are some tips for finding and satisfying BUNS:

  • Turn on “Show bondable atoms” to light up your fold like a jigsaw puzzle of connect-the-dots
  • Show hydrogens (“Show Bondable H”) for an even clearer picture of potential bonding. You can also see this gallery and this table for more details on which amino acids can form what bonds.
  • Struggling to get a sidechain in the right rotation to form a bond? Use the Pick sidechains tool to cycle through rotamers.

Want to get these newsletters directly in your email as soon as they go out? Click here to sign-up and get your weekly Foldit news and tips every Friday!

Until next time, happy folding!

agcohn821's picture
User offline. Last seen 3 hours 51 min ago. Offline
Joined: 11/05/2019
Groups: Foldit Staff
Newsletter October 8: The Pattern Language of Foldit

Hey folders!

Dev Josh here with your weekly Foldit update.

Solutions from This Week's Puzzles

(Disclaimer: This is not scientific feedback; these solutions are not officially endorsed by the Foldit scientists.)

Puzzle 2049: CD47 Binder Design

Puzzle 2050: Symmetric D2 Tetramer Design with AlphaFold Predictions

In Case You Missed It

Want to know more about interpreting AlphaFold structures? Check out this video. Thanks Susume for finding it!

The No BUNS Challenge

Congratulations to this week’s winners!

The challenge will continue through the rest of October, so if you’d like to earn a special Discord role, submit a design solution with the tag “NO BUNS” using the Share with Scientists button.

Today’s Master Folding Tips

I’ve been thinking a lot about design patterns lately. Architecture has patterns to how buildings and urban spaces get designed. So do game design and game programming. But protein design also has design patterns. They’re called motifs, and they’re a big part of bottom-up design. In fact, here’s a whole glossary of protein design motifs. Good design, I’ve learned, is about having a library of references and experiences to draw from. Motifs are good tools to have in your toolkit.

Until next time, happy folding!

agcohn821's picture
User offline. Last seen 3 hours 51 min ago. Offline
Joined: 11/05/2019
Groups: Foldit Staff
Newsletter October 15: The Shake-Wiggle Jiggle

Hey folders!

Dev Josh here with your weekly Foldit update.

Solutions from This Week's Puzzles

(Disclaimer: This is not scientific feedback; these solutions are not officially endorsed by the Foldit scientists.)

Puzzle 2052: IL-2R Binder Design

Puzzle 2053: Symmetric Trimer Design with AlphaFold Predictions

The No BUNS Challenge

Congratulations to this week’s winner: silent gene!

The challenge will continue through the rest of October, so if you’d like to earn a special Discord role, submit a design solution with the tag “NO BUNS” using the Share with Scientists button.

Today’s Master Folding Tips

Let’s talk stabilization. When you want to stabilize your fold, you use shake, wiggle, and clashing importance, right? But in what combination? Generally, you want to start at a low (but non-zero) clashing importance and work your way up to 1.0. Lowering the CI helps your fold pack in well, but lowering it to zero would just make it collapse in on itself. Once your CI is low (around 0.1 or so), give the fold a shake so that the orange sidechains can pull themselves into the core. Then you can wiggle so the backbone can adjust itself, slowly increasing the CI as the protein stabilizes. Sometimes, you might need to go back and forth with shake and wiggle, but usually you want to shake at a low CI so that the sidechains pack in as much as possible. It’s easier to expand a protein than condense it, so start your fold tight and let it relax out.

Want to get the latest Foldit features before everyone else? Check out our devprev update group!

Until next time, happy folding!

agcohn821's picture
User offline. Last seen 3 hours 51 min ago. Offline
Joined: 11/05/2019
Groups: Foldit Staff
Newsletter October 22: Stick It To Those Sheets!

Hey folders!

Dev Josh here with your weekly Foldit update.

Solutions from This Week's Puzzles

(Disclaimer: This is not scientific feedback; these solutions are not officially endorsed by the Foldit scientists.)

Puzzle 2055: CD47 Binder Design

Puzzle 2056: Symmetric Trimer Design with AlphaFold Predictions

Recipe of the Week

When you’re looking to mutate just the interface for a binder design, check out nspc’s mutate interface!

Today’s Master Folding Tips

Having trouble banding sheets together? Turn on stick view! In stick view (with “show bondable H” on), you can more clearly see the hydrogens and hydrogen acceptors: band directly from the hydrogen to its acceptor for great results. Hydrogen bonds form when the hydrogen and its acceptor are around 2 angstroms apart, so set your band length somewhere between 1.75 and 2. Don’t be startled when you switch back to cartoon view and your bands look like they’re in a different place, they’re still where you placed them but they look different in cartoon view.

There’s still one week left in our No BUNS challenge! Submit a design solution with the tag “NO BUNS” using the Share with Scientists button to earn a special Discord role!

Until next time, happy folding!

agcohn821's picture
User offline. Last seen 3 hours 51 min ago. Offline
Joined: 11/05/2019
Groups: Foldit Staff
Newsletter October 29: Let’s Look Longingly at Ligands

Hey folders!

Dev Josh here with your weekly Foldit update.

Solutions from This Week's Puzzles

(Disclaimer: This is not scientific feedback; these solutions are not officially endorsed by the Foldit scientists.)

Puzzle 2058: TGF Receptor Binder Design

Puzzle 2059: VHL Ligand Design

In Case You Missed It

Foldit has released a new Small Molecule Design Tool to accompany the PROTAC small molecule design project.

Today’s Master Folding Tips

This week, we’re all about ligands and small molecules. In a sense, ligand design is all about Foldit fundamentals: pack things together close, but not so close that they clash. Don’t leave any voids, and make sure every acceptor and donor is part of a hydrogen bond. It’s like a jigsaw puzzle where you get to decide the pieces. For the expert designers out there, check out this article on ligand design.

Try out the new tools today in Puzzle 2062: VHL Ligand Design Round 2!

Until next time, happy folding!

agcohn821's picture
User offline. Last seen 3 hours 51 min ago. Offline
Joined: 11/05/2019
Groups: Foldit Staff
Newsletter November 5: A Blueprint for Success

Hey folders!

Dev Josh here with your weekly Foldit update.

Solutions from This Week's Puzzles

(Disclaimer: This is not scientific feedback; these solutions are not officially endorsed by the Foldit scientists.)

Puzzle 2061: Symmetric Trimer Design with AlphaFold Predictions

Puzzle 2062: VHL Ligand Design: Round 2

Today’s Master Folding Tips

Let’s talk about blueprints. If you’re not using the blueprint tool, you should be. Blueprints help you easily sketch out a first draft of a protein and keep your loops ideal, following the Koga & Koga design principles. Here are some of my favorite blueprint blocks:

  • For sheet-to-sheet connections, try the yellow-red loop (including the ends the block shows blue-yellow-red-blue). Yellow-red is good for making a sheet and keeping it flat. If you want your sheets a bit more curved, use the green-green loop instead.
  • For helix-to-helix connections, I recommend the green-blue loop (or red-green-blue-red with the helix ends) for 3-helix bundles. This loop is more common in nature, although you can also use the blue-red-blue loop if you need an extra segment in your loop. For 4-helix bundles, try the green-blue-blue loop because it has a shallower angle. When you’re putting your helices in a binder pocket or making space for a ligand, use the blue or blue-blue loops to angle the helices away and create a space in the middle. Blue-blue in particular is a very wide, almost horizontal angle.

Next week, we’ll cover helix-to-sheet and sheet-to-helix loops!

What are your favorite blueprint blocks? Let me know in the Discord or on the forums!

Until next time, happy folding!

agcohn821's picture
User offline. Last seen 3 hours 51 min ago. Offline
Joined: 11/05/2019
Groups: Foldit Staff
Newsletter November 12: From Sheet To Helix And Back

Hey folders!

Dev Josh here with your weekly Foldit update.

Solutions from This Week's Puzzles

(Disclaimer: This is not scientific feedback; these solutions are not officially endorsed by the Foldit scientists.)

Puzzle 2063: Revisiting Monocyte

Puzzle 2065: Symmetric Trimer Design with AlphaFold Predictions

Today’s Master Folding Tips

This week we’re continuing our tips on using the blueprint tool to make ideal loops.

  • For sheet-to-helix connections, try red-blue (or blue-red-blue-red including the ends) when you have longer helices. This loop works best when your helix length is twice the length of your sheet plus four (for example, with 7-residue strands, use an 18-residue helix) because the outward angle affords longer helices. If you have shorter helices, use the green-blue-blue loop; it points more at a downward angle so you don’t need the helix to be as long to line up with your sheet. Helices in this setup don’t need the extra four residues, so you could use a 14-residue helix for 7-residue strands. Be careful using the blue-blue-blue loop, its angle varies widely in nature so this loop is incredibly unpredictable.
  • For helix-to-sheet connections, the big difference is the number of residues in the loop and whether it angles the sheet toward or away from the helix. Green-blue is a 2-residue loop that points away from the helix; green-blue-red is a 3-residue loop that also points away from the helix; and blue-red-red is a 3-residue loop that points toward the helix. Alternatively, you can think of blue-red-red as a blue-red-red-blue loop, including the first residue of the sheet, which makes it a 4-residue loop that points away from the helix, keeping all three of these loops consistent. In that case, the only difference here is how many residues you need in your loop, and it makes the choice simple.

There’s one blueprint loop I didn’t mention: the blue-red helix-sheet loop. Do you use this loop? Let me know on the Discord when you use it and why!

Until next time, happy folding!

agcohn821's picture
User offline. Last seen 3 hours 51 min ago. Offline
Joined: 11/05/2019
Groups: Foldit Staff
November 19: Check your fold before you wreck your fold!

Hey folders!

Dev Josh here with your weekly Foldit update.

Solutions from This Week's Puzzles

(Disclaimer: This is not scientific feedback; these solutions are not officially endorsed by the Foldit scientists.)

Puzzle 2067: TGF Receptor Binder Design

Puzzle 2068: VHL Ligand Design

Recipe of the Week

If you like to start your symmetric designs by pulling your monomer away from the copies and working on that first, you know how frustrating it can be to find them again. That’s why Bruno Kestemont wrote FindThem, a quick little recipe that helps you pull your symmetric copies back together again and get them in view. Thanks Bruno!

Today’s Master Folding Tips

On puzzles that have AlphaFold predictions, use it early and often to avoid “drilling” into a solution. Sometimes on binder design puzzles, you can make a mutation to your binder that makes it stick better to the target and gives you a better score, but that mutation might make the binder fold up worse on its own. And when scientists try to synthesize your design in the lab, they need it to fold up by itself!

AlphaFold helps because it only predicts how your binder will fold, it doesn’t look at the target. So while you’re working on attaching your binder to the target, keep checking with AlphaFold to make sure your confidence stays high!

Lastly, here’s a pro-tip from nspc: work from the prediction AlphaFold gives you, otherwise AlphaFold’s confidence score might not be reflecting the design you intended to make. This also makes it easier to see changes to the prediction over time as you make mutations.

New to Foldit? Say hi on our Discord! We love new players and we’re always happy to answer questions.

Until next time, happy folding!

agcohn821's picture
User offline. Last seen 3 hours 51 min ago. Offline
Joined: 11/05/2019
Groups: Foldit Staff
Newsletter November 26: Thanks for playing

Hey folders!

Dev Josh here with your weekly Foldit update.

Today is Native American Heritage Day in the U.S., and I’m writing to you from the land of the Oneida people. If you are in the Americas, I encourage you to take a minute today to learn a bit about the indiginous peoples who first lived on these beautiful lands. Yesterday was also Thanksgiving in the U.S., and I am thankful that I get to spend time every week with the Foldit community. All of you are so dedicated to helping science through the puzzles we give you, and every time we are surprised by your ingenuity. So thank you, thank you for playing.

Solutions from This Week's Puzzles
(Disclaimer: This is not scientific feedback; these solutions are not officially endorsed by the Foldit scientists.)

Puzzle 2070: Symmetric Tetramer Design with AlphaFold Predictions

Puzzle 2071: VHL Ligand Design

Today’s Master Folding Tips

Foldit can be fun, but don’t forget to take breaks! It’s good for your body to change positions or stretch every 20 minutes. Check this link for more tips on ergonomic computer usage!

Take it easy this week, folders. Be kind to yourselves, you deserve it.

Until next time, happy folding!

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, Boehringer Ingelheim, RosettaCommons