ligand queue crash is back for 2363

Started by LociOiling

LociOiling Lv 1

On puzzle 2363, at least one of the ligand queue entries causes a crash. The attached log.txt shows the typical crash.

So far, I've been able to collect the details for 28 of the 30 ligands. So there may be two compounds that crash.

(I'm not counting the initial compound when you start the puzzle, the one with weight 411.414. The ligand queue shows a total of 31 entries, but the "Initial" ligand in the queue is always the 411.414 or the last ligand you were working on before the inevitable crash.)

The crash seems to involve this log.txt entry:

interactive.application.actions.ligand_util: {0} Attempting to find the MCS of O=C(Nc1ccc(C(=O)N[C@H]2c3cccc(O)c3C[C@H]2C(=O)O)c(O)c1)NC1CC1 and CNC1=C(F)C=CC=C1Br.Cl

Something about that "Br.Cl" part looks suspicious, but it's not clear what the valid syntax for "find the MCS" is. (Google gives info on solar panel installers.)

log.txt

LociOiling Lv 1

Here's second version of the crash. So it appears there are only 28 non-crashy compounds in the ligand queue.

interactive.application.actions.ligand_util: {0} Attempting to find the MCS of OC[C@@H]1CN(c2nc3cccnc3s2)CC@@HO1 and CNC1=C(F)C=CC=C1Br.Cl

This one has a different entry on the left side of the "find the MCS", so maybe the "Br.Cl" is not the culprit.

log.txt

HuubR Lv 1

According to Wikipedia, the dot in SMILES notation (like in "Br.Cl") is a 'non-bond'. I have no idea why the chlorine would be mentioned in conjunction with this aromatic molecule, when it's not bonded to it. We would have to ask a chemist to explain that.

rosie4loop Lv 1

MCS is probably Maximum Common Substructure, likely when foldit search for similar structures based on the substructure pattern.
BTW a search of Foldit fandom gives a wiki page with a similar message (https://foldit.fandom.com/wiki/Puzzle_1855_reactants)

the dot in SMILES notation (like in "Br.Cl") is a 'non-bond'.

The XX.Cl may represent a salt of the compound, probably need to be removed before 3D conformer generation.

For the SMILES string "OC[C@@H]1CN(c2nc3cccnc3s2)CC@@HO1" pasting it to the Enamine website or smallworld search also giving errors. I am not very familiar with SMILES format, haven't figure out the problem in the SMILES yet.

rmoretti Staff Lv 1

Thanks for the report, and the identification of the compound at issue. It made the solution easy to implement. The puzzle should be updated now. You may need to restart the client in order to get the fix.

The XX.Cl does indeed refer to the chloride salt of the compound. Things apparently weren't properly cleaned prior to submission.

And MCS does stand for Maximum Common Substructure.

OC[C@@H]1CN(c2nc3cccnc3s2)CC@@HO1 is missing square brackets on the second carbon with the chiral hydrogen designation: OC[C@@H]1CN(c2nc3cccnc3s2)C[C@@H]O1 should work – though I'm not sure why that's being annotated as chiral.