Foldit Lab Report #4: 1,994 genes!

Happy New Year! Watch Foldit Lab Report #4 on YouTube now!

- Science Chat recap

- New electron density puzzles (crystals, not cryo)
- We're trying to make poly-proline easier, we promise
- Can you make a Baker lab design even better?
- More and more and more symmetry

- fiendish_ghoul designed a scary good hydrogen-bonded trimer.

- Some of your dimers appear to be working...
- We'll soon be testing almost 2,000 Foldit player designs! We've never been able to generate this much data before. Stay tuned to hear what we learn.

(Wed, 01/01/2020 - 05:19  |  4 comments)
spvincent's picture
User offline. Last seen 6 hours 41 min ago. Offline
Joined: 12/07/2007
Groups: Contenders
tx bkoep. A couple of

tx bkoep. A couple of points:

Everyone has their favourite viewing preferences, but can I suggest that when showing H-bond networks you show the filter? It's easier, for me at least, to see what's going on then.

Also I'd been under the impression that H-bond networks shouldn't include bonds from the sidechain to the backbone. Clearly wrong!

bkoep's picture
User offline. Last seen 1 day 10 hours ago. Offline
Joined: 11/15/2012
Groups: Foldit Staff
Good suggestion!

You're right, the visualization for H-bond Networks definitely highlights the networks more clearly. I'll try to remember to use it in the future!

Yes, sidechains can certainly make H-bonds with the backbone! The way that we've set up the H-bond Network Objective, such H-bonds may not extend your network—but they are still great to have in any design!

Susume's picture
User offline. Last seen 12 hours 15 min ago. Offline
Joined: 10/02/2011
high throughput testing

Thanks for the updates!!

Question about the 1900+ IL-7R binder designs that have been ordered. Is this high-throughput approach available on these puzzles because there is a binder involved? If so, could you say more about how using a binder lets you test lots of different designs at once?

Or, if the binder is not important to testing lots of designs, does that mean we might see high throughput testing on other kinds of foldit designs?

bkoep's picture
User offline. Last seen 1 day 10 hours ago. Offline
Joined: 11/15/2012
Groups: Foldit Staff

To measure binding, we use two techniques called yeast display and fluorescence activated cell sorting (FACS):

In yeast display, we can take a large pool of genes (let's say 10,000), and put them in a yeast culture, such that each yeast cell carries just one of the 10,000 genes. As we grow up the yeast culture, each yeast cell translates its gene into a protein, which is displayed outside the yeast cell membrane. As a result, we have a large heterogeneous population of yeast displaying our 10,000 designed proteins, but in such a way that each yeast cell only displays a single design.

In FACS, we sort out the yeast cells that display successful binder designs. First, we take the target binding protein (e.g. IL-7R), and modify it with a fluorescent chemical "tag." So when we mix this tagged target with our yeast culture, the fluorescent target will stick to any successful binders; any yeast cell displaying a successful binder will be coated with fluorescence. We can then use some very cool microfluidics technology sort out the fluorescent yeast cells (which display a successful design), and finally use DNA sequencing to read back the genes in those yeast cells.

This is a nice experiment because the quantity we measure (fluorescence) is closely linked to the property we want to know (binding affinity); tighter binding should result in brighter fluorescence. If we want to know a different property, then we have to find a way to link that property with fluorescence. For one example of this, you might remember the 2017 paper by Rocklin et al. which used a protease experiment to link fluorescence with folding stability.

User login
Download links:
  Windows    OSX    Linux  
(10.12 or later)

Are you new to Foldit? Click here.

Are you a student? Click here.

Are you an educator? Click here.
Social Media

Only search
Other Games: Mozak
Recommend Foldit
Top New Users

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