1894: Symmetric Trimer Design: H-bond Networks

Closed since over 5 years ago

Summary

• Created: September 24, 2020
• Expires: September 30, 2020 at 23:00 UTC
• Max points: 100

Description

Design a symmetric protein trimer, with 3 identical chains that assemble together! Note the Interaction Energy Objective, which will penalize poorly-scoring large residues that can lead to unwanted protein aggregation. The H-bond Network Objective encourages players to build buried, satisfied H-bond networks at the interface between symmetric chains. H-bond networks are a great way to introduce polar residues at the interface, but it's important that all of the bondable atoms make hydrogen bonds! We've also adjusted the H-bond Network Objective so that poor-scoring H-bonds may not contribute to networks; poor-scoring H-bonds will be displayed in red. This puzzle uses the Buried Unsats Objective, with a large penalty for buried polar atoms that can't make H-bonds. In this puzzle, there are no limits on the Complex Core, but we've included the Complex Core objective so players can see the core residues that can be incorporated into H-bond Networks.

Top groups

• 1. Contenders 100 pts. 21,124
• 2. Gargleblasters 74 pts. 20,899
• 3. Beta Folders 54 pts. 20,783
• 4. Go Science 38 pts. 20,696
• 5. Anthropic Dreams 27 pts. 19,836
• 6. Void Crushers 18 pts. 19,316
• 7. Hold My Beer 12 pts. 18,929
• 8. Marvin's bunch 8 pts. 18,929
• 9. L'Alliance Francophone 5 pts. 18,805
• 10. Russian team 3 pts. 17,084

Top soloists

• 1. spvincent 100 pts. 20,931
• 2. Mike Lewis 96 pts. 20,891
• 3. LociOiling 92 pts. 20,783
• 4. silent gene 88 pts. 20,539
• 5. malphis 84 pts. 20,405
• 6. CharlieFortsConscience 81 pts. 20,353
• 7. ichwilldiesennamen 77 pts. 20,303
• 8. g_b 74 pts. 20,003
• 9. dcrwheeler 70 pts. 19,818
• 10. BootsMcGraw 67 pts. 19,693

Comments

[bkoep]

It's hard to make blanket conclusions about all H-bond networks. But we have definitely noticed a general improvement in the quality of H-bond networks since those changes – especially in the top-ranking solutions.

Before, it was not unusual for us to find a few "unrealistic" H-bond networks (usually with subtle problems in H-bond geometry) among the top-scoring solutions from a puzzle. It seemed these networks were being awarded a big bonus that did not represent a true improvement in design quality.

It makes sense that it's harder to create networks with the recent changes. But it appears that the new HBNet Objective is in fact better about awarding bonuses that reflect real improvements in solutions. (At least, the solutions look better to Foldit scientists; we won't really know about any improvements until we get to test the designs in the lab).

[Formula350]

So I won't ping you over Discord :P
However, we're all getting an error, even with the latest update:

downloading: http://fold.it/portal/files/puzzle_pdb/puzzle_post_2053.zip has key .ocmdline has key .sym has key .roscript TASKOPERATIONS SCOREFXNS FILTERS MOVERS APPLY_TO_POSE PROTOCOLS has key .owts Exception: Both weights file and score function name supplied. in ..\..\..\source\src\standalone\application\boinc\Puzzle.cc line 1403

[bkoep]

Thanks! Should be fixed.

Also, the September 23 update is required to open this puzzle.

[spvincent]

Since the changes to H-Bond Network scoring a few puzzles ago, which were largely aimed at preventing serine triangles and similar constructs, I've found it a lot harder to create networks of any kind. If this means that the new version reflects reality better than so be it, but I can't help wondering if the changes may have been too restrictive.

[bkoep]

It's hard to make blanket conclusions about all H-bond networks. But we have definitely noticed a general improvement in the quality of H-bond networks since those changes – especially in the top-ranking solutions.

Before, it was not unusual for us to find a few "unrealistic" H-bond networks (usually with subtle problems in H-bond geometry) among the top-scoring solutions from a puzzle. It seemed these networks were being awarded a big bonus that did not represent a true improvement in design quality.

It makes sense that it's harder to create networks with the recent changes. But it appears that the new HBNet Objective is in fact better about awarding bonuses that reflect real improvements in solutions. (At least, the solutions look better to Foldit scientists; we won't really know about any improvements until we get to test the designs in the lab).