Hey everyone,
I have been running protein docking with two monomeric proteins by using Rosetta3.2. Everything seemed to be fine with the docking runs except when I added the -dock_ppk line into the flags file.
Instead of having the -dock_ppk line in the flags file, I also tried to prepack the proteins by using the command line:
/usr/local/rosetta3.2/rosetta_source/bin/docking_protocol.linuxgccrelease -database /usr/local/rosetta3.2/rosetta_database/ -s input1.pdb -dock_ppk
It did not work either. Here is the last portion of the log file:
energy_method_options: EnergyMethodOptions::show: etable_type: FA_STANDARD_DEFAULT
EnergyMethodOptions::show: unfolded_energies_type:
protocols.docking.DockingLowRes: EnergyMethodOptions::show: atom_vdw_atom_type_set_name: centroid
protocols.docking.DockingLowRes: EnergyMethodOptions::show: exclude_protein_protein_hack_elec: false
protocols.docking.DockingLowRes: EnergyMethodOptions::show: exclude_DNA_DNA: true
protocols.docking.DockingLowRes: EnergyMethodOptions::show: cst_max_seq_sep: 4294967295
protocols.docking.DockingLowRes: EnergyMethodOptions::show: bond_angle_central_atoms_to_score:
protocols.docking.DockingLowRes: EnergyMethodOptions::show: bond_angle_residue_type_param_set: none
protocols.docking.DockingLowRes: HBondOptions::show: exclude_DNA_DNA: true
protocols.docking.DockingLowRes: HBondOptions::show: use_hb_env_dep: true
protocols.docking.DockingLowRes: HBondOptions::show: use_hb_env_dep_DNA: true
protocols.docking.DockingLowRes: HBondOptions::show: smooth_hb_env_dep: true
protocols.docking.DockingLowRes: HBondOptions::show: decompose_bb_hb_into_pair_energies: false
protocols.docking.DockingLowRes: HBondOptions::show: params_database_tag_: standard_params
protocols.docking.DockingLowRes:
protocols.docking.DockingLowRes: ////////////////////////////////////////////////////////////////////////////////
protocols.TrialMover: Acceptance rate: 0.36
protocols.TrialMover: Acceptance rate: 0.2
protocols.TrialMover: Acceptance rate: 0.36
protocols.TrialMover: Acceptance rate: 0.26
protocols.TrialMover: Acceptance rate: 0.42
protocols.TrialMover: Acceptance rate: 0.6
protocols.TrialMover: Acceptance rate: 0.24
protocols.TrialMover: Acceptance rate: 0.34
protocols.TrialMover: Acceptance rate: 0.36
protocols.TrialMover: Acceptance rate: 0.3
protocols.docking.DockingLowRes: ::::::::::::::::::Centroid Rigid Body Adaptive:::::::::::::::::::
protocols.docking.DockingHighRes: in DockingHighRes.apply
core.scoring.dunbrack: Dunbrack library took 0.02 seconds to load from binary
protocols.docking.DockingHighRes: ::::::::::::::::::DOCK_PPK:::::::::::::::::::
core.pack.task: Packer task: initialize from command line()
Segmentation fault
What could be causing the "segmentation fault"?
Thanks for reading this!
Ideally, Rosetta should give a sensible error message rather than just dieing like that, but segmentation faults generally mean that your system/input files have something about them that the original developers didn't anticipate (but should have).
The best way to trouble shoot them is to compile and run in debug-mode. To compile in debug mode, just leave the "mode=release" off of the scons command (something like "scons -j8 bin/docking_protocol.linuxgccdebug"). To run the debug mode build, use the docking_protocol.linuxgccdebug executable instead of the docking_protocol.linuxgccrelease one.
Debug mode runs much slower than release because it's performing additional checks. Occasionally the debug build by itself will be helpful, but frequently you need to run under a debugger like gdb. To do so, just run "gdb /usr/local/rosetta3.2/rosetta_source/bin/docking_protocol.linuxgccdebug" on the normal command line, and then do a "run -database /usr/local/rosetta3.2/rosetta_database/ -s input1.pdb -dock_ppk" on the gdb command line. After a long while, it'll stop with an error message. You can then run "bt" ("backtrace") to get detailed information on where the program stopped. Usually the error message and the backtrace should be enough to figure out what the problem is. (So you can then quit gdb with the "quit" command.)
Okay, thanks for the suggestion, I will try it a little later.
I forgot to mention that one of the inputing proteins is a model, not a real structure. But the use of the structrual model shouldn't be the cause of the crush becasue I couldn't apply the dock_ppk line to the example from the tutorial either.
The docking_protocol no longer performs prepacking. Use the docking_prepack_protocol executable. This bug has been fixed and will be available in the next release. That is, rosetta will exit gracefully and tell you to run docking_prepack_protocol.
Thank both of you very much for the help. That clears up my doubt/confusion.