- This topic has 4 replies, 2 voices, and was last updated 8 years, 7 months ago by Anonymous.
April 20, 2015 at 7:31 am #2181Anonymous
Hi all, I am facing an error, when reading in the silent output file after a docking simulation. Here I used 6 cores to create 2000 decoys each (sum 12000). See the flag file attached. I used a slightely modified version of the ligand docking application, which is provided in the ROSETTA examples.
Now, i tried to combine all six output files to compare the results. Actually, I tried it with the following command
[path to rosetta]/combine_silent.linuxgccrelease -in:file:silent sil1.out sil2.out sil3.out sil4.out sil5.out sil6.out -out:file:silent fda_dock.out -database [path to database]
with the corresponding paths to the executable and database. The file is read in, but the follwing error is prompted and the program is halted.
core.io.silent: Reading all structures from sil1.out
ERROR:: Exit from: src/core/sequence/AnnotatedSequence.cc line: 116
I also used another approach, because I didn’t know, whether there are different silent file formats. Also, the same error occurs.
[path to rosetta]/score_jd2.linuxgccrelease -in:file:silent sil1.out -out:file:silent sil1_cor.out -out:file:silent_struct_type binary -database [path to database]
Do you have any suggestions, to resolve this.
April 21, 2015 at 8:12 am #10961Anonymous
Unfortunately, all other output files produce another type of error:
core.io.silent: ERROR: trying to index off the end of the secstruct array, idx is 815 secstruct_.size() is 814
terminate called after throwing an instance of ‘utility::excn::EXCN_BadInput’
Abort (core dumped)
During the ROSETTA run, the disc space was nearly empty. I checked the log-files for some error messages, but couldn’t find one. Is there a specific error identifier?
April 24, 2015 at 6:53 pm #10970Anonymous
What type of silent file are you using? Open one of them in a text editor and take a look at the top. If the very top looks like a PDB file, then you’re using an AtomTreeDiff format silent file, and the combine_silent application will not be able to handle it. You’d have to use the score_jd2 application, but pass them in with the -in:file:atom_tree_diff option, rather than -in:file:silent
If that’s not your problem (if the top of your silent file has lines that starts SEQUENCE; SCORE; ANNOTATED_SEQUENCE, etc.) you may want to try adding the -extra_res_fa option with your params file to the combine_silent commandline. Rosetta needs the information about the ligand in order to read in a silent file with that ligand.
The other issue which may be occurring is that your silent files may be corrupt, which is likely given your disk space issues. Try running the commands with the -silent_read_through_errors flag. This will try to skip any corrupted structures, and pull out only the good ones.
April 29, 2015 at 2:00 pm #10989Anonymous
thank you very much for your help. But it worked only partially.
I used binary file type. When I try to combine the output files (4 out of 6 worked) with the flag -silent_read_through_errors then the program abruptly fails with Segmentation fault. There were several erroneous structures, which were skipped (there should be 8000 structures, but only 7947 were read in).
core.io.silent: renamed tag Protein_relA_FDA_2000 to Protein_relA_FDA_2000_3 (SilentStruct with Protein_relA_FDA_2000 already exists!)
core.io.silent: Finished reading 7947 structures from sil6.out
Segmentation fault (core dumped)
I used the following command
combine_silent.linuxgccrelease -in:file:silent sil2.out sil3.out sil4.out sil6.out -out:file:silent fda_dock.out -database /software/linuxsoft/rosetta-3.5/rosetta_database/ -extra_res_fa LG1.params -extra_res_fa LG2.params -silent_read_through_errors
April 30, 2015 at 11:46 pm #10996Anonymous
Would it be possible to recompile in debug mode and re-run? (Just omit the “mode=release” in the scons commandline and use the combine_silent.linuxgccdebug application when re-running. The release and debug compiles can coexist in the same directory.) Often there’s additional error checking in the debug mode which will give (semi) useful error messages in cases when the release mode segfaults.
- You must be logged in to reply to this topic.