Hi there,
I have been trying to submit parameter files (.json) in the “submit file” portal as a part of some tests I am doing, but I think there were some issues in my file and now the run is stuck at 0%. To make sure it wasn’t a server issue I ran another run but manually added the parameters under “Submit” instead of uploading a .json and it is progressing normally. However, I have 4 json files running and they are all stuck and I am getting an error message whenever I try to cancel.
Send us a list of jobiDs
Thanks for the response!
577261
577263
577271
577272
577332
Would you also be able to tell what is the issue with the param files I have written? I am not sure what may be causing the issue. I had written a .py script to edit the .jsons from a run that I made manually, only changing the second protein and the active residue list. For the first four runs I realized that I had not edited the histidine protonation or passive residues list, so i set those to blank and left auto passive and auto his “true” for run 577332, but that one also seems to have gotten stuck.
Do not touch the name of the PDB files (the pdb_file parameter).
Only replace the coordinates when editing the json file.
This is the issue with your modified json files.
I corrected your files and the runs should start again
That fixed it, thanks for your help Dr. Bonvin!
Hi again Dr. Bonvin,
I am once again working on editing my script and seem to have some problem with my parameter file again that I cannot identify. The run moves from queued to running and never makes any progress, but my manual submissions progress just fine. RunID is 585603
Thanks!
your job must have completed as there is nothing on the system for that runID
Yes, the run completed but failed since i sent that last reply. I had another run with more updated settings that completed but failed as well, ID is 585648. Are you able to see completed runs or should I email the outputs?
That run completed but some post-processing failed.
You can download the full archive of the run and look at the generated models.
Is there something in my parameters that is causing these runs to fail during post processing? I have submitted two runs with essentially identical features, the only difference is that I submitted one manually and the other with a .json, but the json version is not progressing and fails at the end. The only difference i made between the two was setting auto_his to true and blanking the his_patch in the json version, but everything else should be identical. RunIDs are 586290 for the json and 586293 for the manual.
Could be simply a glitch…
I would try again and see what happens.
PS: Looks like a rather disordered system you are docking…. I would cut the irrelevant parts.
I am led to believe it is an issue with my file because it is reoccuring with every .json I am generating. I just submitted another pair (one manual, one json) with a different protein being docked and the same behaviour is occuring. IDs are 586765 for the manual and 586758 for the json. It queues and runs no problem, but the it0 progress stays at 0 the whole time. It then fails at around the same time the manual run completes. A lot of the data is there in the failed output but I do also want to access the cluster and graph illustration that the completed runs usually generate.
Then the main question is how do you generate / modify those json files?
If you do a manual submission and it works, use that json for editing but only replace the PBD files (if the molecules are of the same type) and possibly edit the list of active residues
I think i figured out the issue, it had to do with the run directory line of the file. This has led to a new issue for me though. Each time a file is submitted it seems to get a new dir under “pre_process_dir” but the template json I was using has matching locations for preprocess, run, and temptrash dir. However, if I submitted my custom jsons without editing run_dir and temptrash_dir, it is not changed on the new run i submit even though it has a new unique pre_process_dir. I think this has been leading to all my docking runs using the restraints of my template run (which had substrate actives at 206 207 a.a.) even though each custom json has its own set of restraints for unique actives. I’m also not sure how to get around this because the unique preprocess dir isnt generated until i submit the run. Should i just blank the location of run and temptrash before i submit my json, or something else? My other guess had to do with blanking the “tbl_file” line in the json as well, since that contains the restraints as well.
No need to touch any path in the json file.
Only modify the PDB files, restraints and some parameters if needed.