Repository navigation
RosettaDock-5.0 Rosetta processes crashed #773
Replies: 9 comments 4 replies
|
i checked the attached crash log and it contains two distinct failures the docking_protocol reports point to FullatomDisulfideEnergyContainer.cc line 618 in commit 0bdfc78 that index is not included in the crash message you uploaded so the full job stdout or stderr around that line would help identify the affected residue there are also backrub segmentation faults during branch angle optimization could you share the preceding job log and if possible the submitted pdb with the chain identifiers preserved |
|
Thanks a lot.
Which preceding job log should I share with you.
Attachment contains the combined PDB file.
…------------------ Original ------------------
From: ***@***.***>;
Date: Mon, Oct 19, 2026 09:46 AM
To: ***@***.***>;
Cc: ***@***.***>; ***@***.***>;
Subject: Re: [RosettaCommons/rosetta] RosettaDock-5.0 Rosetta processes crashed (Discussion #773)
i checked the attached crash log and it contains two distinct failures
the docking_protocol reports point to FullatomDisulfideEnergyContainer.cc line 618 in commit 0bdfc78
in that exact version the exit occurs when a residue marked as disulfide has no connected partner
the preceding stderr message prints could not find disulfide partner for residue followed by the residue index
that index is not included in the crash message you uploaded so the full job stdout or stderr around that line would help identify the affected residue
there are also backrub segmentation faults during branch angle optimization
those are separate entries so the disulfide message should not be treated as a complete explanation for every crash
could you share the preceding job log and if possible the submitted pdb with the chain identifiers preserved
the crash report alone does not show which input residue or preparation step created the missing connection
—
Reply to this email directly, view it on GitHub, or unsubscribe.
Triage notifications, keep track of coding agent tasks and review pull requests on the go with GitHub Mobile for iOS and Android. Download it today!
You are receiving this because you authored the thread.Message ID: ***@***.***>
|
|
Sorry,
the server does not expose stdout or stderr
…------------------ Original ------------------
From: ***@***.***>;
Date: Mon, Oct 19, 2026 10:00 AM
To: ***@***.***>;
Cc: ***@***.***>; ***@***.***>;
Subject: Re: [RosettaCommons/rosetta] RosettaDock-5.0 Rosetta processes crashed (Discussion #773)
i mean the normal stdout or stderr output from the same failed server job rather than another ROSETTA_CRASH.log or the docking.score file
look for this line in the job output
ERROR: Could not find disulfide partner for residue
please include the residue number at the end and about 20 lines before and after it
in your crash version the program prints that number to stderr immediately before the exit at line 618
that is why the crash report itself has no residue number
if the results page offers a job output download or archive that is the place to check
if the server does not expose stdout or stderr let us know rather than trying to find a file with a guessed name
i also cannot see a pdb attachment in your github reply
it may not have come through with the email reply
please attach the combined pdb directly on the github discussion page or put it in a zip and attach that
without that file or the residue number i cannot identify the broken connection reliably
the exact stderr statement in your rosetta version
—
Reply to this email directly, view it on GitHub, or unsubscribe.
Triage notifications, keep track of coding agent tasks and review pull requests on the go with GitHub Mobile for iOS and Android. Download it today!
You are receiving this because you authored the thread.Message ID: ***@***.***>
|
|
Sorry,
The combined PDB structure has been provided in this previous e-mail.
------------------ Original ------------------
From: ***@***.***>;
Date: Mon, Oct 19, 2026 10:17 AM
To: ***@***.***>;
Cc: ***@***.***>; ***@***.***>;
Subject: Re: [RosettaCommons/rosetta] RosettaDock-5.0 Rosetta processes crashed (Discussion #773)
understood then there is no need to keep looking for stdout or stderr that the server does not provide
the next useful input is the combined pdb itself
please upload it directly in the github discussion as a file or zip because the email reply still has no visible pdb attachment here
with that file we can inspect the chain and residue identifiers and disulfide connectivity records for inconsistencies
the missing partner error is established by the crash location but the exact residue and whether the problem is introduced during server preparation are not established yet
if the submitted file does not reveal the issue the server maintainer would need to inspect the prepared intermediate files and hidden logs using the failed job identifier
we should not guess a residue or remove disulfides just to make the run complete
—
Reply to this email directly, view it on GitHub, or unsubscribe.
Triage notifications, keep track of coding agent tasks and review pull requests on the go with GitHub Mobile for iOS and Android. Download it today!
You are receiving this because you authored the thread.Message ID: ***@***.***>
|
|
I mean the downloadable pdb attachment is visible here in the email replies.
------------------ Original ------------------
From: ***@***.***>;
Date: Mon, Oct 19, 2026 11:37 AM
To: ***@***.***>;
Cc: ***@***.***>; ***@***.***>;
Subject: Re: [RosettaCommons/rosetta] RosettaDock-5.0 Rosetta processes crashed (Discussion #773)
i checked the github version of both email replies and there is no downloadable pdb attachment visible there
+
+please open this discussion in your browser and drag a zip containing the combined pdb into a new comment
+wait for the upload to finish before posting
+
+if the structure cannot be shared publicly send it through the server support channel with the failed job id
+we need the actual file or the server prepared structure to identify the residue rather than guessing from the crash report
—
Reply to this email directly, view it on GitHub, or unsubscribe.
Triage notifications, keep track of coding agent tasks and review pull requests on the go with GitHub Mobile for iOS and Android. Download it today!
You are receiving this because you authored the thread.Message ID: ***@***.***>
|
|
Sorry, due to the restrictions of the non-disclosure agreement, I cannot paste the download link to the pdb in the discussion or upload the zip using the github website.
------------------ Original ------------------
From: ***@***.***>;
Date: Mon, Oct 19, 2026 11:48 AM
To: ***@***.***>;
Cc: ***@***.***>; ***@***.***>;
Subject: Re: [RosettaCommons/rosetta] RosettaDock-5.0 Rosetta processes crashed (Discussion #773)
understood the attachment may be visible in your email copy
i am reading the github discussion rather than your email inbox
the only downloadable file currently visible on the github page is ROSETTA_CRASH.log
could you paste a download link to the pdb in the discussion or upload the zip using the github website
then i can inspect the actual structure
—
Reply to this email directly, view it on GitHub, or unsubscribe.
Triage notifications, keep track of coding agent tasks and review pull requests on the go with GitHub Mobile for iOS and Android. Download it today!
You are receiving this because you authored the thread.Message ID: ***@***.***>
|
|
Using ROSIE admin access, I was able to take a peek at the files for one of your failing jobs(*). The main issue you're running into is the backrub segfault (the first error in the file). That Segfault is causing the backrub protocol to crash hard, and in doing so it can mess up the rest of the protocol. (That's why you're not getting results for some of your runs -- the backrub crash doesn't allow any of the downstream processing to properly execute.) I can't 100% confirm, but I think the first backrub error you're getting is due to the same issue as #756 -- One thing to try would be to flip the order of things, such that the antigen is the receptor and the antibody is the ligand. Regarding the disulfide error, the relevant file is *) While private jobs aren't visible to the general web audience, we do not and can not make any guarantees about the absolute security of files submitted to ROSIE. If you have data which requires a stringent level of access control, please do no submit to the public ROSIE server. You'll be better off running things locally. (Additionally, some bugs on the ROSIE server may have been fixed in the most recent released or development version of Rosetta.) |
|
Thanks a lot.
I will try to flip the order of things, such that the antigen is the receptor and the antibody is the ligand.
------------------ Original ------------------
From: ***@***.***>;
Date: Tue, Oct 20, 2026 05:04 AM
To: ***@***.***>;
Cc: ***@***.***>; ***@***.***>;
Subject: Re: [RosettaCommons/rosetta] RosettaDock-5.0 Rosetta processes crashed (Discussion #773)
Using ROSIE admin access, I was able to take a peek at the files for one of your failing jobs(*).
The main issue you're running into is the backrub segfault (the first error in the file). That Segfault is causing the backrub protocol to crash hard, and in doing so it can mess up the rest of the protocol. (That's why you're not getting results for some of your runs -- the backrub crash doesn't allow any of the downstream processing to properly execute.)
I can't 100% confirm, but I think the first backrub error you're getting is due to the same issue as #756 -- One thing to try would be to flip the order of things, such that the antigen is the receptor and the antibody is the ligand.
Regarding the disulfide error, the relevant file is 5-docking.log. This indicates that it's residue H 22 which is at issue, but that should be recognized as being disulfide bonded to H 96. Everything looks reasonable as far as inputs. The one thing I can possibly see is that one of the backrubbed ensemble members is missing that disulfide for some reason. I'm not sure why. Ideally the docking ensemble protocols would be robust to these sorts of errors, but it might be choking on the fact that the main input has the disulfide where the ensemble member does not. That said, it looks like the docking_protocol application is robust to this -- if it encounters the error, it will simply skip that output and move on to the next. That's why you can see some results for JLDOCK3, despite the ROSETTA_CRASH.log. -- I'd only worry about this error if you're not getting sufficient output results.
*) While private jobs aren't visible to the general web audience, we do not and can not make any guarantees about the absolute security of files submitted to ROSIE. If you have data which requires a stringent level of access control, please do no submit to the public ROSIE server. You'll be better off running things locally. (Additionally, some bugs on the ROSIE server may have been fixed in the most recent released or development version of Rosetta.)
—
Reply to this email directly, view it on GitHub, or unsubscribe.
Triage notifications, keep track of coding agent tasks and review pull requests on the go with GitHub Mobile for iOS and Android. Download it today!
You are receiving this because you authored the thread.Message ID: ***@***.***>
|
|
Thanks a lot.
I try to flip the order of complex structure, the antigen is the receptor and the antibody is the ligand. Although it appears that
ERROR: Some of the Rosetta processes crashed!
, the docking PDBs were presented in the output results.
Best wishes.
Lun, jiang
------------------ Original ------------------
From: "Rocco ***@***.***>;
Date: Tue, Oct 20, 2026 05:04 AM
To: ***@***.***>;
Cc: ***@***.***>; ***@***.***>;
Subject: Re: [RosettaCommons/rosetta] RosettaDock-5.0 Rosetta processes crashed (Discussion #773)
Using ROSIE admin access, I was able to take a peek at the files for one of your failing jobs(*).
The main issue you're running into is the backrub segfault (the first error in the file). That Segfault is causing the backrub protocol to crash hard, and in doing so it can mess up the rest of the protocol. (That's why you're not getting results for some of your runs -- the backrub crash doesn't allow any of the downstream processing to properly execute.)
I can't 100% confirm, but I think the first backrub error you're getting is due to the same issue as #756 -- One thing to try would be to flip the order of things, such that the antigen is the receptor and the antibody is the ligand.
Regarding the disulfide error, the relevant file is 5-docking.log. This indicates that it's residue H 22 which is at issue, but that should be recognized as being disulfide bonded to H 96. Everything looks reasonable as far as inputs. The one thing I can possibly see is that one of the backrubbed ensemble members is missing that disulfide for some reason. I'm not sure why. Ideally the docking ensemble protocols would be robust to these sorts of errors, but it might be choking on the fact that the main input has the disulfide where the ensemble member does not. That said, it looks like the docking_protocol application is robust to this -- if it encounters the error, it will simply skip that output and move on to the next. That's why you can see some results for JLDOCK3, despite the ROSETTA_CRASH.log. -- I'd only worry about this error if you're not getting sufficient output results.
*) While private jobs aren't visible to the general web audience, we do not and can not make any guarantees about the absolute security of files submitted to ROSIE. If you have data which requires a stringent level of access control, please do no submit to the public ROSIE server. You'll be better off running things locally. (Additionally, some bugs on the ROSIE server may have been fixed in the most recent released or development version of Rosetta.)
—
Reply to this email directly, view it on GitHub, or unsubscribe.
Triage notifications, keep track of coding agent tasks and review pull requests on the go with GitHub Mobile for iOS and Android. Download it today!
You are receiving this because you authored the thread.Message ID: ***@***.***>
|
Uh oh!
There was an error while loading. Please reload this page.
Dear everyone, I recently used online RosettaDock for protein-protein docking. I submitted antibody-antigen combined PDB files. The results showed that
ERROR: Some of the Rosetta processes crashed!
Usually this indicates issues with the job input. Please see log files for details.
ROSETTA_CRASH.log
All reactions