Sidechains - Why Bother with Them?

It’s easy to ignore Sidechains in Foldit; you can get decent scores without ever looking at them and just playing with the Backbone's shape. But sidechains are what make each protein fold the way it does. The backbone is chemically identical in all proteins and serves only as scaffolding. It's the sidechains' interactions with their watery environment that causes the protein to bend and fold up the way it does.

So, even though we tend to focus on the shape of the backbone in Foldit, moving it around and rebuilding and so on, it’s good to remember that it’s the sidechains we’re trying to make happy - we want to minimize sidechains clashing with each other, and minimize stress in the backbone angles caused by sidechain quibbles. Foldit gives us some great Tools to do this. You just need a little Strategy to get the best possible scores.

How to Look at SidechainsEdit

First off, there are two ways to look at sidechains in Foldit. They’re a bit confusing at first, but quite helpful once you figure them out.

1. Score mode. Control-Shift-S puts you in Score mode (Command-Shift-S for Macs). From the View menu, check the boxes for these options: "Glow hydrophobic sidechains," "Use relative score coloring," and " Show All (slow)." Now you can see all the sidechains in various colors. In Score Mode, the base of the sidechain glows blue if it’s hydrophilic and orange if it’s hydrophobic. The color of the body of the sidechain (and the backbone) tells you how “happy” that sidechain is. Blue is happy, gray/brown is bored, orange is unhappy, and red needs anger management classes.

2. Hydrophobic Mode. Access the other view by hitting Control-Shift-H (Command-Shift-H for Macs). ( I think this stands for Hydrophobic Mode – I never did find these commands in HELP). In this view, the sidechain body is blue if it’s hydrophilic and orange if it’s hydrophobic. The base of the sidechain will glow to indicate its happiness. Again, blue is happy, red is unhappy, and no color means somewhere in between.

You’ve probably noticed that both views give you the same information – reversed. I find that the Hydrophobic Mode is cleaner for working with sidechains, and the Score Mode is better for assessing trouble spots from a backbone perspective.

Playing with SidechainsEdit

You should already know how to use Global Shake and Local Shake; if not, see Shake Sidechains. But you need to learn how to deal with sidechains individually to really get things going. Here are the moves you need to know:

1. How to Shake one sidechain.

  • Zoom in on an area of interest.
  • Move your cursor over the sidechain you want to move.
  • That sidechain should “light up”, meaning it’s ready to be moved. (In any view but “Show All (Slow) you have to hover over the base of the sidechain.)
  • Double click the sidechain.
  • It will shake back and forth, looking for a better local position. It generally doesn’t find one if you already done Global/Local Shakes. More on how to use it below.

2. How to manually move one sidechain.

  • Do the same steps as above to highlight a chain.
  • Drag to move the sidechain around. It helps to drag the end of the sidechain, not the base.
  • You’ll notice that the sidechain will only stop in a few positions. That’s normal. Just keep waving it around, and you’ll see what positions are possible.
  • Release the drag to choose a position.

3. Learn how to freeze a sidechain.

The Most Basic Sidechain StrategyEdit

1. Choose a random sidechain and manually move it to a new position. Go ahead, cause a clash.

2. Global Wiggle and see what happens.

Not very sophisticated, but you can get sometimes results with seemingly random moves like this.

Walking the SidechainsEdit

Start at one end of the protein and work your way to the other end, adjusting the position of each sidechain once or twice, then wiggling to see what happens. Most of the time the wiggle will result in a lower score BUT the backbone may be in a new alignment that will respond to a shake and move other sidechains into higher scoring positions.

How to Pick a SidechainEdit

You can get better results by choosing sidechains that are more likely to give up points. First, focus on areas that are unhappy (red or orange). Try lots of configurations; finding a better position for one of these sidechains can yield big points (after wiggling actions). If you don't get much out from moving unhappy sidechains, focus on the large ones, even if they're relatively "happy." Side chains to try would be the ones with hexagons/pentagons and anything long or flexible. Even seemingly random moves can give big point increases with enough patience.

Ring StackingEdit

When amino acids with ring structures (the hexagon-looking sidechains) are close together in the center of the molecule, these rings generally prefer to be parallel (more or less) with each other – not at right angles to each other. This is especially true for phenylalanine (the hexagon *without* the stick at the end) and the others to a lesser degree. Not only does "stacking" help compact the protein, but the rings interact with each other and stabilize the protein. So try moving ring sidechains into a more stacked or parallel configuration.

You can find out more about ring stacking in this article.

The Explosion StrategyEdit

Find a juicy sidechain and move it to cause a big clash - like your score goes to "0," for example. Now, Global Wiggle. After your protein explodes and starts to come together again, stop the Wiggle. Double click your juicy sidechain; its position will probably change. From here, I've gone in two directions, each with occasional results. First, I Wiggle again and let the protein come to a full rest. Then I Shake and Wiggle / Nudge if the score is close to my previous. Second, I sometimes Wiggle just a little more, then stop and Shake that one sidechain again, then Wiggle to a rest, and do the usual Shake, Wiggle.

There's no one correct way to do it. My theory is that the clash causes the protein to fly apart in a different way than pulling, sometimes allowing it to find a more favorable configuration as it settles down. This method works in late Middle Game (or early End Game??) - when the protein's structure is pretty well defined and it wants to return to its configuration after being pulled apart back together, but before any refinement has been done (like slice/dice).

A More Refined StrategyEdit

You need your powers of observation for this one. Look closely at an unhappy sidechain and its surroundings and ask yourself, “Why is Sidechain A unhappy? How could I make it happier?” You don’t need chemistry for this, just the basics you’ve already learned. For example, it might be unhappy because:

  • it’s too close to a neighbor;
  • it’s hydrophobic but is facing outside the protein;
  • its position is causing strain in the backbone;
  • its ring structure wants to be more parallel with its neighbors.

Based on your theory of why it's unhappy, move the sidechain to a position you think it might like better. Your score will plummet; don’t worry, that’s normal. Now look around to see what clashes you caused that made your score plummet. Try to fix those clashes. Here are some methods:

  • Leave Sidechain A alone, and try to fix those other clashes by individually shaking the sidechains. Then Global/Local Wiggle, Nudge, etc.
  • Lock Sidechain A, then Global Wiggle (or Local Wiggle), unlock, Global Wiggle again, Nudge, etc.
  • Lock Sidechain A, and Global Shake (or Local Shake). Global Wiggle, etc.

There's no set order to do things in; try different combinations. If you move one sidechain and your first Wiggle or Shake gets you close to your original score (maybe with 10 points or so), you might be onto something. Keep Wiggling and Nudging; you can often outdo your best score.