Merge pull request #2035 from RosettaCommons/aleaverfay/movemaps_from_residue_selectors
This PR will introduce a new MoveMapFactory class that is more-or-less equivalent to a TaskFactory:
It will construct a MoveMap (equiv: PackerTask) just ahead of MoveMap use given a Pose
It holds instructions (equiv TaskOperations) for specifying the MoveMap behavior using ResidueSelectors and (new!) JumpSelectors
This will add a MoveMapFactoryDataLoader and a JumpSelectorDataLoader into RosettaScripts to define new top-level elements.
The MoveMapFactory begs for an additional TorsionSelector or DOF_Selector mechanism for selecting DOFs that are otherwise inaccessible. That'll go on a to-do list for now.
Perhaps with the MoveMapFactory and with ResidueSelectors, we have enough justification for sweeping through existing Movers and Filters in an XRW trying to get them using this friendlier interface for controlling behavior, as proposed at WRC by @cdbahl
notify author
notify list [rosetta-logs@googlegroups.com]
Merge pull request #2068 from RosettaCommons/everyday847/swa_accelerations
Make SWA faster in two ways
Rocco Moretti 8 years This commit broke the LoopModeler unit tests, as well as the build.header test (which is ultimately the cause of the PyRosetta test failures).
Andy Watkins 8 years I'm currently rebuilding -- will fix!
notify author
notify list [rosetta-logs@googlegroups.com]
Merge pull request #2069 from RosettaCommons/revert-2029-roccomoretti/ubsan_fixes
Revert "Fix integration.ubsan"
Reverts RosettaCommons/main#2029
Caused major build issues. Reverting until things are fixed.
notify author
notify list [rosetta-logs@googlegroups.com]
Merge pull request #2029 from RosettaCommons/roccomoretti/ubsan_fixes
Fix a few issues which were causing issues in the integration.ubsan tests:
* There's a bug in boost::format that invokes undefined behavior: https://svn.boost.org/trac/boost/ticket/11632
* There was some uninitialized variables in GlycanRelaxMover
* For the thermal_sampler integration test, the params_ variable of RECCES_Mover isn't getting set. I don't know enough about this to say where it should be getting set - instead I just added checks such that similar oversights will be much more visible. The thermal_sampler integration test is now on perma-failure until the issue is addressed.
notify author
notify list [rosetta-logs@googlegroups.com]
Merge pull request #2067 from RosettaCommons/vmullig/tp1_integration_test
Add integration test for protocol used in upcoming paper.
I'm not 100% sure that everything used is already covered by tests; nor am I sure that the integration of the components will continue to work in the future.
Tasks:
- Add the protocol (used to design threefold-symmetric peptide 60mers with threefold-symmetric crosslinkers).
- Strip it down so that it runs in ~1 min but still tests the essential workflow.
Rocco Moretti 8 years Turns out the log_* files from the threefold_symm_peptide_design integration test are *not* deterministic - they include the SHA1 of the code. -- With recent master, the ignore_list should be robust to MPI-style output, so I'd recommend just using that.
Vikram K. Mulligan 8 years Yes, I saw that. I've got a pull request open in which I fixed this in my branch.
notify author
notify list [rosetta-logs@googlegroups.com]
Merge pull request #2049 from RosettaCommons/jadolfbr/skip_loading
CYD backwards compatibility
Replace CYD with CYS:disulfide in the name_map of ResidueTypeSet for backwards compatibility of features databases (among others).
notify author
notify list [rosetta-logs@googlegroups.com]