puzzle 2363 --compounds 385.445 and 385.552 don't get CL bonus

Started by LociOiling

LociOiling Lv 1

On puzzle 2363, the compounds with weight 385.445 and 385.552 aren't earning the compound library bonus.

Since the bonus is 4000 points in this round, these two compounds seem like non-starters.

Both compounds have 1a and 1b entries that show "Similarity 1.000000". Neither shows the "Exact Match" displayed for other compounds.

For 385.445 and 385.552, having the compound library window open doesn't earn the bonus. Accepting the a or b version also doesn't earn the bonus. Shaking and wiggling doesn't help in any of these cases.

Loading a saved solution doesn't earn the bonus, even with the "Load Library Compound" window already open.

LociOiling Lv 1

Add the compound with weight 390.417 to the "no bonus" list. It has 1a, 1b, and 1c versions.

Only 385.445, 385.552, and 390.417 from the initial ligand queue show this problem.

This problem seems to happen when there are multiple versions (isomers?) for a compound library match, at least in some cases.

So in the example above, for 385.552, entries 2a and 2b also don't get a bonus when you click "Accept Compound".

If you scroll down to entry 3, there's no 3a and no 3b. Accepting entry 3 does get the bonus.

On the other hand, from the initial compound 411.414, if I load result 25a for compound 368.449, I do get the bonus. So the lack of bonus doesn't apply to every case where there are multiple versions of a compound.

But working from 411.414 again, loading result 1a for compound 385.445 still doesn't earn the bonus.

rmoretti Staff Lv 1

Could you try this again, with a fresh submission to the compound library? I tweaked some things about the library search, and I think that might have fixed things.

LociOiling Lv 1

Compound library searches are just hanging at the moment.

In one case, after just restarting Foldit, I now see the bonus for 385.445. But then on a second freshly started Foldit, I don't see the bonus. Both clients are on devprev, the only difference I can see is that I changed tracks on the one that doesn't show bonus.

At the moment, 385.552 and 390.417 still don't receive the bonus, but who knows. New searches for these two are hanging.

Also, I've noticed in some cases, there will be a 1a with "Exact Match", then 1b and so on with "Similarity 1.000000". The bonus gets awarded in that situation. It's only when there's no "Exact Match" shown that the problem crops up.

rmoretti Staff Lv 1

We're having some disk issues with the system which runs the compound library search. It theoretically shouldn't have cause things to go down, but apparently there's been some knock-on effects. Things should (hopefully) be back up sometime today.

Regarding restarting clients, remember that you'll need to download the updated compound libraries (open the Compound Library panel) and potentially perform an action (e.g. wiggle) in order to get the compound library bonus to show up.

Having 1a and 1b with "Exact Match" and "Similarity 1.000000" is the intended behavior. The various letter-specified entries from the compound library arise because chemical suppliers don't always specify the full enatiomeric/chiral state (often because its easier to just sell a mix of them), but in order to actually model them in 3D you need to pick one. The Compound Library creates the alphabetic variants if this is the case, but only one of them will match the exact 3D enantiomer you have in the structure. (Though if you pull from the Ligand Queue or Compound Library, there should be an Exact Match when you re-submit it.)

LociOiling Lv 1

All my pending searches from yesterday completed. New searches are completing quickly.

Compounds 385.552 and 390.417 still don't get the compound library bonus, even with a new search submitted and completed today.

Both still show only "Similarity 1.000000" matches, no "Exact Match" entries.

Update: after a restart, compound 390.417 gets the bonus. Odd, since that client had only been running to 10 minutes. Perhaps it has something to do with the compound already being loaded when the client starts.

Still no bonus for compound 385.552, even after a second restart. Having the compound pre-loaded doesn't seem to have helped.

LociOiling Lv 1

Compound 390.417 seems consistently inconsistent. Generally, restarting Foldit gets the compound library bonus. A restart may not work 100% of the time, however.

Compound 385.552 is still not getting the bonus, despite restarts, fresh searches, shakes, wiggles, and so on.