Lab Report #17: New data from the lab!

We have some exciting news to share after testing some of your protein designs in the lab! Plus we've got new puzzles and a fresh Design of the Month.

- Symmetric designs and small angle x-ray scattering (SAXS) experiments

- More designable linker puzzles
- Symmetric tetramer design

- A symmetric design, by silent gene, with a substantial core that spans the entire protein. See for yourself in the new Design of the Month sandbox puzzle.

(Mon, 02/01/2021 - 14:23  |  4 comments)
spvincent's picture
User offline. Last seen 24 min 24 sec ago. Offline
Joined: 12/07/2007
Groups: Contenders

tx bkoep: interesting as always.

I was hoping you could expand on what "rigid" means in the context of the linker puzzles. Is one big helix rigid enough? Maybe a 3-helix bundle would be better. Or perhaps many-stranded sheets would be good. How about relatively long loops with lots of prolines in them?

bkoep's picture
User offline. Last seen 54 min 58 sec ago. Offline
Joined: 11/15/2012
Groups: Foldit Staff
Great question!

Foldit isn't built to model protein dynamics, so it takes a little bit of imagination to think about how your protein might be able to flex or change shape.

I tend to think about hydrophobic packing as the glue that holds the protein together. If two parts of the protein (like a helix and strand) are in contact via orange hydrophobic residues, then I think of them as stable and rigid; if you tried to pull the helix and strand apart from one another, then you would expose the hydrophobics to water and the pulled-apart structure would be unstable.

On the other hand, if two regions are only in contact via blue polar sidechains (making H-bonds, for example), then you could pull them apart relatively easily. The polar sidechains, which used to be making H-bonds with each other, can now make H-bonds with surrounding water, so they are stable either way.

When I try to imagine the "rigidity" of a design, I look for parts of the protein that can be pulled apart in this way. Suppose I have two well-folded helical bundles that are connected by a loop; if the two bundles do not pack against one another with hydrophobic residues, then the connecting loop will behave like a hinge that allows the bundles to flop around relative to one another.

You can also look for long stretches of only-blue polar residues. These regions are liable to unravel, since they are not held in place by any hydrophobic packing. A single long "megahelix" is not very rigid for this reason; if it is not glued to anything, then the all-blue stretches of the helix will readily unravel.

Prolines are less flexible relative to other residues, true. But a proline-rich region will still need some hydrophobic packing in order to lock it down into a rigid fold. In fact, in our protein designs we should really try to avoid any kind of loop that is longer than a few residues, if we can help it.

neilpg628's picture
User offline. Last seen 23 hours 19 min ago. Offline
Joined: 08/01/2019
Groups: Foldit Staff

Hi Spvincent,

In general rigid means that the linker will hold the binders in their preset orientation. This would increase the binding affinity of the combined complex.

For the context of Foldit that would mean that the linker is well-folded, but also has secondary structure elements that connect the termini of the linker with the termini of the binders. If the connections between the binder termini and the linker termini are floppy then the linker can be well folded without preserving orientation.

This is also why packing against the binders themselves is rewarded.

spvincent's picture
User offline. Last seen 24 min 24 sec ago. Offline
Joined: 12/07/2007
Groups: Contenders

Thanks for the responses: most helpful.

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, RosettaCommons