saved 1294 solution crashes client
| Case number: | 845829-2002942 |
| Topic: | Crash/Hang |
| Opened by: | LociOiling |
| Status: | Open |
| Type: | Bug |
| Opened on: | Saturday, October 15, 2016 - 04:47 |
| Last modified: | Sunday, October 16, 2016 - 07:55 |
A saved solution for 1294 with 78 residues crashes the client when loaded.
The crash generates a debug.txt containing the following error:
"message" : "Error in core::conformation::Conformation::residue(): The sequence position requested was greater than the number of residues in the pose."
The solution was generated on 20161006-39881adc28-win_x86-devprev.
As a workaround, adding the two missing residues and saving again allows the new solution to be loaded. The extra residues can then be deleted. It's difficult to avoid damaging the pose when adding residues, simply adding and deleting residues results in a loss of score. Even freezing the entire protein before adding segments does not prevent score loss.
Crashing and non-crashing solutions shared with scientists.
Workaround found. The "later share" mentioned in the last post opened successfully for me and teammates. But then the same share caused a crash, as did teammates' subsequent shares. Somewhat baffling.
The workaround is to load a solution with the full complement of 80 residues, then delete residues to match the number in the share you want to open. This allows the "short" share with 78 residues to be opened. The initial extended chain can be used, just delete the surplus residues, and your short shares load correctly.
I realized that my earlier success with the "later share" happened because I loaded it on top of solutions that had already been reduced to 78 residues. Clients which had been working on 80-residue solutions crashed when loading the 78-residue share.
A later share from the same client loads for me without a problem. No need to add missing residues.