I'm following this tutorial to run a protein structure prediction: https://new.rosettacommons.org/demos/latest/tutorials/rosetta_cm/rosetta_cm_tutorial
I was able to model one protein sequence successfully following those steps, but trying for different ones I've been getting this same error:
ERROR: Assertion `tlen < slen` failed.
ERROR:: Exit from: src/core/fragment/util.cc line: 177
protocols.rosetta_scripts.ParsedProtocol: [ ERROR ] Exception while processing protocol:
File: src/core/fragment/util.cc:177
[ ERROR ] UtilityExitException
ERROR: Assertion `tlen < slen` failed.
protocols.jd2.JobDistributor: [ ERROR ]
[ERROR] Exception caught by JobDistributor for job S_0001
[ ERROR ]: Caught exception:
File: src/core/fragment/util.cc:177
[ ERROR ] UtilityExitException
ERROR: Assertion `tlen < slen` failed.
It repeats other 2 times for each job (S_0002 and S_0003). I checked to see if the files were appropriately edited with the right pdb/fasta/grinshin/frags files, and couldn't find any problems.
I will be attaching the crash log to this post. Any help would be much appreciated!
| Attachment | Size |
|---|---|
 ROSETTA_CRASH.log | 56.81 KB |
Category:
Post Situation:
I'm not quite sure what's going on, but it looks like there's an issue with your 3-mer fragment files. I'd double check that you've generated those correctly, and there isn't any corruption in the files.
I'll mention that with recent versions of Rosetta, you might not need fragments anymore for RosettaCM. There's an internal fragment generation process. The fragments aren't as high of quality as you get with a dedicated fragment picker stage, but if you're just trying to close loops and patch over short insertions/deletions (rather than trying to build domains, which RosettaCM isn't good at anyway), they're sufficient.
I'll also plug the Meiler Lab tutorial for RosettaCM, which I think has been updated a bit more recently than the demo you link: https://meilerlab.org/rosetta-workshop-2022/
Thank you so much for your answer, really appreciate you sending the updated tutorial!
Just figured out the problem, my fragment files were in a DOS format, not Unix. Formatting them with ":set fileformat=unix" with vim solved the problem!