1436: NMN Binder Design: Round 2

Closed since over 8 years ago

Summary

• Created: October 02, 2017
• Expires: October 10, 2017 at 23:00 UTC
• Max points: 100

Description

This is a followup to Puzzle 1422. Redesign the loops of this protein to better bind the ligand! Nicotinamide mononucleotide (NMN) is a small-molecule that makes up part of the small-molecule nicotinamide adenine dinucleotide (NAD), which is sometimes regarded as 'the molecule of youth.' Scientists know that the amount of NAD in our cells declines as we age, and so designing a protein to bind to NMN is the first step towards achieving protein-based biosensors that can detect NMN levels in cells. Such biosensors would help scientists to understand the aging process and could potentially lead to a 'cure' for aging! In this puzzle, players have a little more freedom in redesigning the NMN binding pocket. Players may load in solutions from Puzzle 1422.



The starting structure of this puzzle is a protein designed to bind NMN, but we think it can be improved. There are several loops near the NMN binding pocket that could be extended to make stronger interactions with the ligand. In the selected loops, players may insert and design up to 10 extra residues for free; additional residues may be added at a cost of 15 points per residue. Most of the protein backbone is frozen, and there are strong constraints to keep the ligand positioned in the binding pocket. There are permanent cutpoints flanking each loop, so that players may freely manipulate the loops without disturbing the rest of the structure.

Top groups

• 1. Gargleblasters 100 pts. 8,360
• 2. Go Science 80 pts. 8,357
• 3. Marvin's bunch 63 pts. 8,329
• 4. Anthropic Dreams 49 pts. 8,329
• 5. Contenders 37 pts. 8,231
• 6. Beta Folders 28 pts. 8,201
• 7. Void Crushers 21 pts. 8,163
• 8. L'Alliance Francophone 15 pts. 8,135
• 9. HMT heritage 11 pts. 7,922
• 10. :) 8 pts. 7,890

Top soloists

• 1. Enzyme 100 pts. 8,360
• 2. ZeroLeak7 98 pts. 8,351
• 3. anthion 95 pts. 8,323
• 4. actiasluna 92 pts. 8,319
• 5. frood66 90 pts. 8,318
• 6. Bruno Kestemont 87 pts. 8,298
• 7. Susume 84 pts. 8,277
• 8. isaksson 82 pts. 8,253
• 9. crpainter 80 pts. 8,225
• 10. Galaxie 77 pts. 8,206

Comments

[toshiue]

client scoreboards not updating, have heard the same from three other group members.

[Susume]

From the H-bond network puzzles we learned the idea of bondable atoms being "satisfied" if they have H bonds. I assume that idea also holds in the ligand puzzle (in theory, if not in the score).

I have a choice of histidine or lysine in one location to bond to a red atom of the ligand. Foldit gives better score to the histidine, but it removed a threonine that I had put in to "satisfy" the red atom of the histidine, so now I have a mostly-buried red atom that is "unsatisfied." If I put the lysine in, its one bondable atom bonds to the ligand, but it's only one bond - seems like sometimes lysine forms up to 3 H bonds. In sphere view, neither the HIS nor the LYS are very accessible to water.

Is a histidine satisfied if its blue atom is H bonded and its red one is not?
Is a lysine satisfied if its blue atom has only one H bond?

Foldit doesn't seem to penalize unsatisfied buried atoms or to reward satisfied ones, and it's frustrating trying to guess what will work. Is there some way the score (or some other feature) could be used to give us feedback on this? It seems like it will always be an issue in ligand puzzles.