Placeholder image of a protein
Icon representing a puzzle

2136: KLHDC2 ligand design: Round 1

Closed since 7 months ago

Intermediate Overall Small Molecule Design


April 20, 2022
Max points

Design a small molecule which can bind to KLHDC2! We don't want you to modify the protein, but instead use the small molecule design tools to build new small molecules which fit into the KLHDC2 binding pocket.

Note: To get the most out of the small molecule design tools, we recommend changing you view settings to the Small Molecule Design Preset.

The starting ligand is a peptide which binds to KLHDC2. We're interested in designs which make it less peptide-like and with better TPSA, but which keep the carboxylate interaction and the hydrophobic pocket interaction. See the blog post for more background information, and pay attention to the various objectives. Note that as the KLHDC2 domain is a bit large, we've trimmed the protein down to just the pocket (this shouldn't affect the prediction of compound binding interactions).

This is a project in collaboration with Boehringer Ingelheim & Oxford University. Boehringer Ingelheim & Oxford have committed to help evaluate and test the molecules which Foldit players have designed. The structures of all compounds created as part of the collaboration puzzles as well as any experimental results from testing them will be made publicly available. All participants and game sponsors of current and future small molecule design games commit to complying with the Foldit Terms of Service including those pertaining to intellectual property.

Top groups

  1. Avatar for Contenders 100 pts. 22,831
  2. Avatar for BIchallenge 2. BIchallenge 70 pts. 22,787
  3. Avatar for Anthropic Dreams 3. Anthropic Dreams 47 pts. 22,641
  4. Avatar for Gargleblasters 4. Gargleblasters 30 pts. 22,330
  5. Avatar for Go Science 5. Go Science 19 pts. 21,719
  6. Avatar for Australia 6. Australia 11 pts. 21,450
  7. Avatar for Void Crushers 7. Void Crushers 7 pts. 20,747
  8. Avatar for Marvin's bunch 8. Marvin's bunch 4 pts. 20,678
  9. Avatar for L'Alliance Francophone 9. L'Alliance Francophone 2 pts. 20,124
  10. Avatar for Foldit Staff 10. Foldit Staff 1 pt. 19,988

  1. Avatar for NeLikomSheet
    1. NeLikomSheet Lv 1
    100 pts. 23,021
  2. Avatar for Aubade01 2. Aubade01 Lv 1 95 pts. 22,894
  3. Avatar for spvincent 3. spvincent Lv 1 90 pts. 22,794
  4. Avatar for julianfuchs 4. julianfuchs Lv 1 84 pts. 22,787
  5. Avatar for BootsMcGraw 5. BootsMcGraw Lv 1 80 pts. 22,656
  6. Avatar for gmn 6. gmn Lv 1 75 pts. 22,640
  7. Avatar for Bletchley Park 7. Bletchley Park Lv 1 70 pts. 22,613
  8. Avatar for robgee 8. robgee Lv 1 66 pts. 22,506
  9. Avatar for vo218 9. vo218 Lv 1 62 pts. 22,465
  10. Avatar for infjamc 10. infjamc Lv 1 58 pts. 22,335


rmoretti Staff Lv 1

There's a number of objectives. Most of them are there to keep the molecule going "too far" outside the range of acceptable parameters for drugs.

All objectives give +1000 points when fully satisfied, for a maximum of +12 000. Most objectives have a "Show" button which should highlight areas of concern in red. Even when fully satisfied, objectives may highlight groups which may eventually contribute, so it's not recommended to keep Show on at all times.

  • Carboxylate Interaction - A bonus for keeping the hydrogen bonds the existing carboxylate is making. (Show highlights the protein atoms you need to make the hydrogen bonds to.)
  • Ligand Hydrophobic Interaction - A bonus for keeping a hydrophobic interaction between the ligand and the protein. (Show highlights the protein residues the ligand should contact.)
  • Peptide - Penalizes the formation of peptide-backbone-like groups. (Show highlights which parts of the ligand should be changed.)
  • Torsion Quality - Keeps bond rotations in a good range. Using Wiggle or Tweak Ligand can fix bad torsions. (Show highlights torsions to be rotated.)
  • Number of Rotatable Bonds - Intended to keep the ligand from getting too big and floppy. You can reduce rotatable bonds by deleting groups or forming rings. (Show higlights rotatable bonds.)
  • Ligand TPSA - Topological Polar Surface Area - Keeps the polar surface area (including buried polar surface) low. To improve, try removing oxygens and nitrogens. (Show highlights atoms contributing to higher TPSA.)
  • Ligand cLogP - A measure of polarity - Keeps the molecule from getting too hydrophobic. To improve, try adding polar oxygens and nitrogens. (Show highlights atoms contributing to higher cLogP.)
  • Ligand Hydrogen Bond Donors - Keep the number of ligand hbond donors low. (Show highlights atoms contributing to higher cLogP.)
  • Bad Groups - Gives a bonus for avoiding groups that aren't very drug-like, including enol, enamine, thiocarbonyl, hydroxylamines, geminal diols, alpha-beta unsaturated carbonyls, aliphatic alkanes, aliphatic alkenes, aliphatic ethers and others. (Show highlights groups at issue.)
  • Element Counts - Limits certain element numbers: Desired P=0, S<=1, I<=1, Br<=1, Cl<=4, F<=4, N<=5 (Show highlights atoms of elements exceeding count.)
  • Molecular Weight - Intended to keep the ligand from getting too big.
  • Synthetic Accessibility - Identify molecules which are likely to be hard to make in the lab. (Show highlights regions of the molecule contributing to difficult synthesis.)

LociOiling Lv 1

The protein part of this puzzle has been chopped up into many chains. Some of the chains are only three or four residues (amino acids) long.

The protein in question is PDB 6DO5, titled "KLHDC2 ubiquitin ligase in complex with USP1 C-end degron".

The protein is also known as "Kelch domain-containing protein 2", which is where the KLHDC2 comes from.

The full protein is 363 residues, so cutting it down to 96 residues seems like a good idea. (The full protein also forms a dimer, so there would be two of them to contend with….)

A few Foldit recipes are chain aware, and look at the protein part as chains A through O, with the ligand as chain N. The dialogs for print protein, AA Edit, and SS Edit need a little tweaking to handle 16 chains, unless you have a high resolution monitor in portrait mode. Fixes will be forthcoming.

Sandrix72 Lv 1

Unfortunately this new puzzle has many bugs. For example adding a C atom killed the program completely. I have to restart it.

LociOiling Lv 1

I'm seeing the same kind of crashes adding an atom or a fragment.

After a crash, if you're on Discord, you can drag the log.txt file from c:\Foldit (or the equivalent spot) over to the #bugs-and-feedback channel. Sometimes log.txt can contain what appears to be useful information. Other times, not so much.

The drag-and-drop trick works if you're using the Discord app. I'm not sure how you do it with the web version, but I assume there's a way.

rmoretti Staff Lv 1

From the crash reports, we've identified the likely cause of the crashing, and we have a fix which we think should address it which hopefully should be coming out soon.

rmoretti Staff Lv 1

The fixes which should address the crashing issues should now be in both main and devprev.