Merge pull request #76 from RosettaCommons/nollikai/CoupledMover Adding CoupledMover to protocols/simple_moves Created a new mover that performs coupled side chain / backbone moves. For ligands, this mover performs a rigid body rotation and translation coupled with ligand rotamer optimization. This mover currently uses ShortBackrubMover for backbone moves, BoltzmannRotamerMover for side chain and ligand rotamer moves and RigidBodyPerturberMover for rigid body moves. This mover is compatible with PackerTasks that allow positions to be designed, making it easy to incorporate backbone flexibility into sequence design. Since ligand moves can also be performed, this mover is particularly useful for the design of ligand binding sites. In the near future I will commit a protocol that uses this mover to predict the set of tolerated sequences for a given design problem by keeping track of all the sequences that are accepted during a Monte Carlo simulation.

• Summary

fixing tests (My text editor erased two needed spaces again; grr.)

• Summary

Carbohydrates: mostly database updates This merge includes some addition of ATOM_ALIASes, a new sugar topology file, (so that I can load in sialyl-Lewis X,) and some sugar pilot app updates to reflect the recent options system clean-up. It also adds a formal +1 charge to all lower terminus peptide residues. ---------------------------------------- Unit test status: Not Run (no C++ code changed) Integration test status: 1 test change carbohydrates: expected, due to the +1 on peptide lower termini

• Summary

Adding PyRosetta unit tests

• Summary

Merge pull request #75 from RosettaCommons/nollikai/ClashBasedRepackShellSelector Adding ClashBasedRepackShellSelector to core/pack/task/residue_selector Created a new ResidueSelector that identifies residues that clash with any possible rotamer of all designable positions in a given PackerTask. This solves a problem I routinely encounter with distance-based repack shells, which never include the same residues I would have chosen from looking at the structure. Instead, this ResidueSelector can be used to define a repack shell that only includes residues whose side-chains clash with any rotamer of a design position, resulting in the selection of residues that I actually want to be repacked. This code was co-written with Roland Pache @rpache

• Summary

Atom Typing: fixing two bugs related to atom typing NEWS: The following bugs have been fixed: - The patch system failed to correctly set chemical::Atom's MMAtomType name, so so any variant ResidueType had mismatched MMAtomType names and indices. - The patch system failed to correctly set new Elements when an AtomType changed. This was primarily an issue for cases when virtual atoms were changed to real atoms and vice versa. - The Gasteiger atom-typer failed to correctly assign GasteigerAtomTypes in those cases, because it relies on having the correct Element. (As an aside, I have also add a miscellany of my own glyco-PyRosetta pilot apps for tracking in Git.) ---------------------------------------- Unit test status: Pass Integration test status: 1 test changed carbohydrates: now showing correct atom typing

• Summary

Merge pull request #74 from RosettaCommons/dougrenfrew/normalmode_warning_fix Fix warning-as-errors error in NormalMode code

• Summary

Merge pull request #73 from RosettaCommons/nollikai/ShortBackrubMover Adding ShortBackrubMover to protocols/simple_moves Created a new mover that performs a three residue backrub move. Peptide bond rotations are also performed to minimize the displacement of backbone carbonyl oxygen atoms. This mover samples rotation angles from a normal distribution, unlike traditional backrub moves, which use a uniform distribution. The standard deviation of the rotation angle distribution can be set using set_rotation_std_dev( Real ). This mover is particularly useful when combined with BoltzmannRotamerMover to perform coupled side-chain / backbone moves.

• Summary

Merge pull request #72 from RosettaCommons/PyRosetta Fixing PyRosetta namespace build

• Summary

Eliminating CircularPermutationFinder filter. In retrospect the filter didn't make much sense. No test changes expected.

• Summary

ResidueProperties: fixing unit test; sorry! :rat:

• Summary

ResidueProperties: VariantTypes are now enums instead of strings NEWS: Similar to what I did with properties, **`VariantType`s are now `enum`s and the variant state of a given `ResidueType` is now stored in a vector of Booleans indexed by said `enum`s.** This is Part IIA of my refactoring of `ResidueType` properties in Rosetta. **Done:** - [x] Variant status stored in `ResidueProperties` as `vector1< bool >` and indexed by `VariantType`. - [x] Auto-code-generation script renamed and refactored to generate multiple files. - [x] String-look-ups removed from ~170 files. - [x] ~90 topology and patch files modified to reflect name-changes to `VariantType`s that would conflict with properties. - [x] Added new unit tests. - [x] Fixed any broken unit tests. - [x] Fixed any broken integration tests. - [x] Added way to handle on-the-fly `VariantType`s. (IMO, they should be avoided, but it is allowed here for backward compatibility.) - [x] Moved variant matching functions out of `ResidueType` and into `util.cc` and thoroughly refactored them. **Still To-Do:** - [ ] Refactor more central code that still relies on string-look-ups. (Much of the code involving switching from one variant to another involves making a list of all the `VariantType` strings in the before and after states and comparing them to select an appropriate `ResidueType`. I will fix this shortly.) Most of the "Still To-Do" is complete or nearly complete. I am trying to confirm that numeric changes in SWA and SWM integration tests are not something about which to worry. I thus decided to break this merge into two parts. If you are curious as to the progress, see pull request #56 on GitHub. ---------------------------------------- Unit test status: Pass Integration test status: 4 tests change inverse_rotamer_remodel, enzdes, Enzrevert_xml: output changes explaining custom VariantTypes carbohydrates: output changes to ResidueType::show()

• Summary

Fixing copy constructor in BoltzmannRotamerMover Setting randomize_resnum_ to true if resnum_ is not set.

• Summary

Trying to fix what I broke in the last commit after changing the src/protocols/cluster/cluster.cc file. I added a new flag to allow clustering without superimposition during distance measure calculation.

• Summary

Made some changes to cluster.cc file. Added a flag called -skip_align to allow clustering without superimposition during RMSD matrix calculation. Used rmsd_no_super for that purpose.

• Summary

Removing unused variables in BoltzmannRotamerMover to get rid of warnings

• Summary

Legend: