Let's Foldit! - Episode 1 of the Foldit podcast
Happy Wednesday everyone! We're proud to present the first episode of our new podcast all about Foldit and the science behind it. In this episode we sit down with Brandon Kier, one of the scientists working on Foldit, to discuss his latest puzzle: 675 Hydrogen-producing Catalyst Dimer.
Be sure to check out Brandon's companion blog post as well!
A transcript is available below.
Attachment | Size |
---|---|
Ep1Transcription.doc | 39.5 KB |

why do ya'll want negative design?
Negative design here refers to making an alternative, unwanted potential conformation not work. It's tricky to do. Ideally, we'd show you two different structures: your design bound to the catalyst, and your design bound to an alternate version of the catalyst. (We've frozen the catalyst for the puzzle, but in real life it can wiggle, kind-of like protein sidechains, and when it wiggles into another conformation it doesn't work as well.) That way, you could compare how your design looked for the correct form of the catalyst with how it looked when attached to the wrong form; ideally, the wrong form would look horrible (lots of clashes, voids, etc.) while the correct form would look great - and this would help fix the catalyst in the proper orientation.
For now, we've only shown you the desired, catalytically active "good" conformation. The concept of negative design can be potentially confusing to players ("wait, you want me to make a really BAD structure?") and tricky to set up in a Foldit puzzle, since we'd have to show two structures (desired vs. bad) at once.
I'll probably take some of the top-scoring structures and attach them to the "wrong" catalyst form on my own, to make sure that they'll prefer the "good" catalyst, but I'm not sure whether we'll be able to make a specific Foldit puzzle for it. Though it would be really cool! I'll keep bugging the devs about it. :)
It would be possible to implement a visual version of the negative design with the puzzle types that already exist in Foldit. If the incorrect catalyst was shown as a guide, in the same manner as a QTTN puzzle, then at least it would be possible to visually detect clashes, or more specifically to fold the protein to deliberately clash with the guide, assuming that the difference between the two catalysts was visually significant.
A good suggestion. We actually were discussing this option when we were setting up the puzzle. It would likely work quite well visually.
Thanks for podcast Kate and Brandon. Looking forward to more (=