Merge pull request #142 from RosettaCommons/nollikai/coupled_moves_ligand_weight
Fixing missing input file in coupled_moves integration test
This change should allow the coupled_moves integration test to run properly.
notify author
notify list [rosetta-logs@googlegroups.com]
Add header that is missing, but curiously doesn't trigger missing errors with gcc/clang
Replacing line core/pack/interaction_graph/SurfaceInteractionGraph.hh:1957
core::scoring::TenANeighborGraph const & tenA_neighbor_graph( get_surface_owner()->pose().energies().tenA_neighbor_graph() );
With the exploded form
SurfaceInteractionGraph< V, E, G >* surf_owner_ptr( get_surface_owner() );
assert( surf_owner_ptr );
SurfaceInteractionGraph< V, E, G > const & surf_owner( *surf_owner_ptr );
core::pose::Pose const & debug_pose( surf_owner.pose() );
core::scoring::Energies const & debug_energies( debug_pose.energies() );
core::scoring::TenANeighborGraph const & tenA_neighbor_graph( debug_energies.tenA_neighbor_graph() );
results in a compiler error about core::scoring::Energies being an incomplete type, even though the single line one doesn't. Adding the Energies header fixes it.
notify author
notify list [rosetta-logs@googlegroups.com]
Merge pull request #95 from RosettaCommons/kylebarlow/general_dev
Change JD2 to throw exception if a job fails
The prior behavior was for JD2 to exit cleanly and return a zero exit code, even if the only job it tried to run failed. This means that integration tests can sometimes fail silently when they completely fail to run.
JD2 will now count failed jobs (from bad inputs of FAIL_DO_NOT_RETRY), and if this count is above 0, throw a specialized exception at the end of go_main(), which would typically then be caught by the application and result in a non-zero exit code.
I have modified the JD2 integration test to fail against the old case, but it now should still pass.
Integration tests that appear to be broken, and which will now fail continuously after this commit:
coupled_moves
loop_creation
filter_termini_near_interface
antibody_graft
splice_out_shorter_tail
No failed unit tests expected.
notify author
notify list [rosetta-logs@googlegroups.com]
Merge pull request #140 from RosettaCommons/phbradley/atomtype_mods
atomtype mods: new command line mechanisms to modify atom type chemical properties from the command line. -clone_atom_types for cloning pre-existing atomtypes; -reassign_atom_types for changing the atomtype of specificied atoms in a ResidueType
notify author
notify list [rosetta-logs@googlegroups.com]
Merge pull request #137 from RosettaCommons/nollikai/coupled_moves_ligand_weight
Adding option to weight protein-ligand interactions in coupled_moves protocol
The weight of residue-ligand interactions can now be specified using the flag coupled_moves::ligand_weight (Real). The weight is 1 by default, meaning that residue-residue and residue-ligand interactions are scored equally. Setting ligand_weight to 2 up-weights the residue-ligand interaction energies by a factor of 2.
This branch also contains two other changes, including changing the access specifier of set_last_accepted_pose and set_lowest_score_pose in MonteCarlo to public and adding compute_one_and_two_body_energies to RotamerSet. These changes were needed to incorporate ligand weighting into BoltzmannRotamerMover and the coupled_moves protocol.
notify author
notify list [rosetta-logs@googlegroups.com]
Merge pull request #128 from RosettaCommons/rpache/clash-based_repack_shell_pilot_app
adding a pymol string to the output of the create_clash-based_repack_shell pilot app
notify author
notify list [rosetta-logs@googlegroups.com]
Merge pull request #131 from RosettaCommons/rhiju/change_swm_rna_o2prime_phosphate_packing_defaults
Rhiju/change swm rna o2prime phosphate packing defaults
Tests did finish here:
<a href="http://benchmark.graylab.jhu.edu/revision?id=380&branch=">http://benchmark.graylab.jhu.edu/revision?id=380&branch=</a>
This is a really minor update in some flags, and highly isolated to stepwise code. All integration test changes in integration.release are expected, except a few weird things remarked upon below.
@gideonla @sarelf Actually I'm not sure about splice_out_tail: http://benchmark.graylab.jhu.edu/sub_test/6409795
You might have an uninitialized variable -- some boolean or Size in a class that is either not initialized or cloned properly. I tracked one issue down previously in the splice_ tests, but won't have time to do the same here...
@aleaverfay No clue why match_1ct2 keeps failing. The diffed lines look the same to me: <a href="http://benchmark.graylab.jhu.edu/sub_test/6409884">http://benchmark.graylab.jhu.edu/sub_test/6409884</a>.
@rfalford12 some of the score tests (like score_jd2) are showing erratic behavior in the display of 'total_score' vs. 'score' columns in .fsc files. Only seems to show up in integration.release_debug, not integration.release comparisons. Probably an uninitialized variable -- I didn't track down fully but saw that you recently made some updates to SilentFileJobOutputter. Let me know if you can't find the fix!
notify author
notify list [rosetta-logs@googlegroups.com]