Merge pull request #709 from RosettaCommons/JWLabonte/sugars/loops
Patch System: adding new PatchOperation DELETE_VARIANT_TYPE
This merge allows one to remove a VariantType from a ResidueType, expanding all the fun things we can do with the patch system. I also needed it to make modified sialic acids, which are key to eukaryotic glycobiology.
also added a new integration test to the carbohydrates tests to demonstrate
carbohydrates test changes expected
notify author
notify list [rosetta-logs@googlegroups.com]
Merge pull request #694 from RosettaCommons/JWLabonte/sugars/database
Carbohydrates: adding more sugars to database
This merge simply adds some more sugars and a fancy bacterial modification patch to the database.
In the process, I fixed tracers in `PatchOperation.cc`, which affects output in two integration tests that have the tracer volume jacked up. (Why they do, I do not know.)
notify author
notify list [rosetta-logs@googlegroups.com]
Merge pull request #704 from RosettaCommons/rhiju/fix_patch_selectors_nucleobase_sample_around
turn on patch selectors that are needed in nucleobase_sample_around. …
The problem tests are the three 'usual' broken integration tests & mac.pyrosetta (also now broken).
notify author
notify list [rosetta-logs@googlegroups.com]
Merge pull request #706 from RosettaCommons/sergey/f
Adjusting CMake config: adding CXX11 build, making default Clang build on Mac to use -std=c++11 to match our scons settings.
notify author
notify list [rosetta-logs@googlegroups.com]
Merge pull request #701 from RosettaCommons/weitzner/dock_partners_flag
Dock partners in any order
====================
This change set enables docking chains to be specified in any order. The previous functionality required that the dock_partners flag listed the chains in the same order as in the PDB file. For example, if you had a PDB file with the chains A, B and C, you could dock the A–B dimer to C ("AB_C") or A to the B–C dimer ("A_BC"), but not the A–C dimer to B ("AC_B"). To use "AC_B", the user would have to rearrange the chains in the PDB file. Absurd! :angry:
This change set fixes that using `ChainSelectors` and representing dock partners as `vector1< bool >`s. There were many flavors of the `setup_foldtree` function, several of which were not called anywhere in the code. These functions have been removed with the author's permission as appropriate.
Unit tests:
-------------
All unit tests pass. Unit tests that tested functions that were removed were also removed and the `DockingProtocolFunctions.u` file was updated to account for `FoldTree` edges being printed in a different order. A new unit test has been added to demonstrate and test the out-of-order functionality.
Integration test changes:
---------------------------------
Seventeen integration tests differ (3 were failing on the branched revision, 2 are failing as a result of an unknown bug somewhere else) as a result of this change, but the differences are limited to the order in which `FoldTree` `Edge`s are printed out and do not affect the folding direction of the pose or cause trajectory changes.
### Tests with text-based differences ###
These differences are all of the form:
< protocols.docking.DockingProtocol: FOLD_TREE EDGE 1 71 -1 EDGE 71 107 -1 EDGE 71 126 1 EDGE 126 193 -1 EDGE 126 108 -1
---
> protocols.docking.DockingProtocol: FOLD_TREE EDGE 1 71 -1 EDGE 71 107 -1 EDGE 71 126 1 EDGE 126 108 -1 EDGE 126 193 -1
1. `docking_local_refine_min`
2. `docking_low_res`
3. `score_only_silence`
4. `fuzzy`
5. `assemble_domains_jd2`
6. `docking_local_refine`
7. `rotamer_probability`
8. `docking_distance_constraints`
9. `docking_site_constraints`
10. `docking_full_protocol`
11. `score12_docking`
12. `docking_ensemble`
### Tests that were failing on the merge base with master:
1. `place_simultaneously`
2. `hotspot_graft`
3. `inverse_rotamer_remodel`
### New test failures: ###
1. `mp_dock_prepack` (does not run) – ***disabled with permission from JKL***
2. `mp_find_interface` (does not run) – ***disabled with permission from JKL***
notify author
notify list [rosetta-logs@googlegroups.com]
Merge pull request #702 from RosettaCommons/JWLabonte/sugars/input
Carbohydrates: making Rosetta correct PDB files created by GLYCAM on input
GLYCAM is a popular MD tool for carbohydrates. Unfortunately, it has its own three-letter code and treats the HO- group or RO- group of reducing end residues as their own residues. This merge will try and have Rosetta fix this problem automatically with the -glycam_pdb_format flag.
This merge will also make it so that one no longer has to use the -override_rsd_type_limit with sugars. -include_sugars will set that option automatically.
only expected carbohydrates test changes from changed flag and input files
notify author
notify list [rosetta-logs@googlegroups.com]
Merge pull request #648 from RosettaCommons/jkleman/mp_updates
Jkleman/mp updates
SASA:
* added functions in core/scoring/sasa/util that (1) compute the total per-residue-SASA, (2) relative per-residue-SASA, and (3) is_residue_exposed boolean function; this is all based on Jared's code
* added sasa calculator pilot app that uses those functions: pilot/jkleman/per_residue_sc_sasa.cc
* added integration test
Apps:
* add helix_from_sequence app with integration test, works both for soluble as well as membrane helices
* cleaned up interface_statistics app for membrane proteins, still more to refine
Docking:
* MPDockingSetupMover: added options to optimize membrane positions for both partners before concatenation
* MPFindInterfaceMover: cleaned up, added options, checking in the code from before Gordon conference, still needs work
* MPQuickRelaxMover: cleaned up, added options, checking in code from before Gordon conference, still needs work
RosettaMP:
* MembraneInfo: added function in_membrane that returns boolean
* SpanningTopology: fixed bug in concatenating topology objects that was not caught by unit test, included that in unit test now
* Embedding: added functions for add_span_embedding
* TransformIntoMembraneMover: added option to optimize embedding, still needs test
* TranslationRotationMover: removed unused fullatom variable and option that was never used
* VisualizeEmbeddingMover: debugged, but more debugging to do
* most membrane movers: added check at the beginning of the apply function to check the foldtree, after apply the foldtree is reset to the original one
* fixed a whole bunch of bugs, unit test oversights and integration test changes that weren't caught in the renaming of the integration tests for the membrane framework
* cleaned up some of the center/normal tags and commandline initializations that didn't yet use Rebecca's function
* protocols/membrane/util: added a whole bunch of useful functions: create_membrane_foldtree with anchor points at chain COMs, create_membrane_foldtree with anchor points at COMs of TM regions of chains, helper function for both, chain COM, chain TM COM, rsd closest to chain COM, rsd closest to chain TM COM, split topology by chain, compute embeddings by chain; added unit tests for all of those
Other:
* pose/util: added get_chains function
Unit test failures: none
Integration test changes: mostly cosmetic, helix_from_sequence and per_res_sc_sasa didn't exist, mp_mutate_relax and mp_quickrelax changed because the protocol changed
notify author
notify list [rosetta-logs@googlegroups.com]
Merge pull request #703 from RosettaCommons/vmullig/pull_request_demo
Updating a comment in a pilot app.
This is to demonstrate how to make a pull request.
You can make lists:
Like this.
notify author
notify list [rosetta-logs@googlegroups.com]
Merge pull request #695 from RosettaCommons/rfalford12/remove_old_rm_code
Removing old resource manager hooks in the membrane framework
----------------------------------------------------------------------------------------------
When the membrane framework was in development, we added hooks to the resource manager. However, development stopped on these. Currently, the code does not compile, and if it did, it would not work with the current version of the RosettaMP framework. Once we update to JD3/resource manager, I'll try this again but probably from scratch. No reason for non-compiling, and non-functional files to complicate master.
notify author
notify list [rosetta-logs@googlegroups.com]