Merge pull request #346 from RosettaCommons/vmullig/mutate_bug
vmullig/mutate_bug
Trying to fix an irritating bug in the MutateResidue mover and elsewhere.
When a residue is mutated to another type, the default behaviour was for any atoms with matching names to have their positions preserved, and for any missing atoms to be rebuilt from ideal coordinates. This was fine for Y->F or F->Y mutations, but created serious problems when dissimilar residues had similar atom names. The new default behaviour will be to copy mainchain heavyatom positions only, with everything else built from ideal coordinates; chi values will then be copied from the old residue to the new. The old behaviour can be restored with a RosettaScripts tag when setting up the MutateResidues mover.
Tasks completed / TODO:
Overloaded copy_residue_coordinates_and_rebuild_missing_atoms() to have a version that takes a bool. False yields old behaviour, true yields new.
Set up new behaviour (rebuilding all but mainchain heavyatoms, copying chi values).
Created unit test for new behaviour.
Modified MutateResidue mover so that its settings can be set without invoking parse_my_tag() by functions called from parse_my_tag().
Updated MutateResidue mover and added flag to restore old behaviour.
Updated Remodel mover (really, the form_disulfide() function in core/conformation/util.cc).
Updated MutateResidue documentation.
Expected integration test changes due to this merge:
--remodel_disulfides_rosettascripts: Expected and desired change. The added disulfides now have beta-carbon protons in the right places.
--unfolded_state_energy_calc: Looking at the code, this calls the MutateResidue mover, so this is also expected. (The change is small.)
--msd_mover: The FindConsensusSequence mover that this test's script calls uses the MutateResidue mover, so this is also an expected change. (The change is very small.)
notify author
notify list [rosetta-logs@googlegroups.com]
Merge pull request #345 from RosettaCommons/nollikai/coupled_moves_output_structures
Added flag to coupled_moves protocol to save structures for all designs
notify author
notify list [rosetta-logs@googlegroups.com]
Merge pull request #347 from RosettaCommons/vmullig/bgs_parallel
Silly off-by-one error in BundleGridSampler.cc -- fixed.
Apparently I can't count. Bundle grid sampler nstruct mode integration test change is the only expected test change. I'm going to merge this without testing.
notify author
notify list [rosetta-logs@googlegroups.com]
Merge pull request #324 from RosettaCommons/roccomoretti/braaaiiins
File-level dead code removal
Remove files that aren't compiled or used by any of the standard processes (regular build, unit test, PyRosetta) This includes both .cc files and header files. Only the main library directories (basic, utility, numeric, core, protocols) have been cleaned. "External" libraries (I include ObjexxFCL in this) weren't touched. I also didn't touch the unit test directory (I want to make it easy to re-enable those tests which aren't running currently), nor the apps or devel directory (as I didn't want to penalize people who sped up compilation by turning off their devel work). -- Though if we're sunsetting the devel directory, a cleanup there may be warranted.
If I deleted something you needed, feel free to add it back (remember it's still available in git's history), but: 1) make sure you put the .cc file in the appropriate .src.settings file 2) make (and enable) a unit test for the functionality
No test changes expected, as the removed code shouldn't be used.
notify author
notify list [rosetta-logs@googlegroups.com]
updating the site.settings.tacc.stampede for compatibility with new pointer system. These setting have been tested on stampede and work properly as of this commit. Thanks to Cyrus Proctor for helping me determine the correct settings
notify author
notify list [rosetta-logs@googlegroups.com]
Merge pull request #340 from RosettaCommons/revert-328-JWLabonte/sugars/input
Revert "Properties/VariantTypes Refactor Part IIB & CarbohydrateInfo Pro...
(I didn't notice that I had changed one integration test. Oops!)
notify author
notify list [rosetta-logs@googlegroups.com]
Merge pull request #328 from RosettaCommons/JWLabonte/sugars/input
Properties/VariantTypes Refactor Part IIB & CarbohydrateInfo Property-Handling Refactor
The purpose of this code is to ultimately allow one to generate poses with branched and/or modified carbohydrates from annotated sequences or `.GWS` files. Currently, branching and modified sugar residues can only work if loaded in from a `.pdb`.
To do this, I realized I needed to go back to my unfinished refactor of the `Properties`/`VariantTypes` system just a bit, so this merge will include a mixture of changes.
Please see: `source/src/core/chemical/residue_properties/README.txt` in this pull request for what I'm doing with the `Properties`/`VariantTypes`.
All unit tests pass.
Carbohydrate integration test changes are expected.
notify author
notify list [rosetta-logs@googlegroups.com]
Merge pull request #313 from RosettaCommons/vmullig/genkic_sampling
Create a bin transition probability-based backbone sampling method.
This allows sampling based on the probability of residue i being in mainchain torsion bin X, given the mainchain torsion bin of one or both adjacent residues.
Organization/Tasks:
In database/protocol_data/generalizedKIC/bin_params/, added data files for bin transition probabilities.
In core.3 (core/scoring/bin_transitions/) created a new BinTransitionCalculator class with methods for calculating bin transition probabilities, and a new BinTransitionData class for storing bin transition probability matrices and derived data. These classes also have methods for reading in bin transition probability data files. (Note that this has to be in core.3 because it draws on functions from the Ramachandran class, and it will be used by energy methods among other things.)
Added a unit test for some of the core functionality of the above.
In protocols/generalized_kinematic_closure/perturber/GeneralizedKICperturber class, added a new "randomize_backbone_by_bins" perturber.
In protocols/generalized_kinematic_closure/perturber/GeneralizedKICperturber class, added a new "perturb_backbone_by_bins" perturber.
Adding an integration test for GenKIC with bin-based initialization.
Adding an integration test for GenKIC with bin-based perturbations.
Adding a stand-alone randomize mover.
Adding a stand-alone perturb mover.
Adding integration tests for the stand-alone randomize mover.
Adding integration tests for the stand-alone perturb mover.
Documentation.
Put off to a future pull request:
Figure out how to do the stats for D-amino acids, and add D-amino acids probability matrices to the database.
notify author
notify list [rosetta-logs@googlegroups.com]
Merge pull request #333 from RosettaCommons/vmullig/bluegene_mpi_problem
Vmullig/bluegene mpi problem
On the Argonne "Mira" Blue Gene/Q machine, I was having a problem with the MPI version of the RosettaScripts app. Every time a slave process asked for a job, the master process would send it a nonzero job number, but the slave process would interpret it as "0". This fixes that problem, at least with the MPIWorkPoolJobDistributor.
In MPIWorkPoolJobDistributor, core::Size values were being sent and received as MPI_INTs. On little-endian systems (e.g. x86 processors), the data truncation resulted in the value being preserved for any reasonably small value, which is why, I think, no one had seen the problem anywhere else, but on big-endian systems (like Blue Gene), for a small value like a job number, only the leading zeroes were sent and received. I switched all the MPI_INTs to MPI_UNSIGNED_LONGs (which should match core::Size on most systems -- core::Size is actually of type std::size_t, which is usually unsigned long), and adjusted a few ints defined locally to be core::Sizes. This got RosettaScripts working properly in MPI mode on Blue Gene.
Note that MPI_Send and MPI_Recv take a virtual pointer to the datum being sent or recieved, so there is no check that its type matches the type that MPI thinks it's sending or receiving. Data truncation errors are easy with MPI.
notify author
notify list [rosetta-logs@googlegroups.com]
Merge pull request #100 from RosettaCommons/kwikwag/peptiderive-talaris
A refactor of PeptideDeriver, aka Peptiderive.
It was a pilot app. Now it's a protocol + app.
**Main classes and features:**
- `PeptideDeriverFilter` (`protocols/analysis`, `protocols.3`) - the essence of the protocol. Given a pose, calculates the peptidic stretch contributing more than others to the interaction energy. It also reports on and optionally models cyclizable peptides (thanks to @orlypolo ). Data is written down to the log or to a file, in one of two formats.
- `DisulfideInserterMover` (`protocols/simple_moves`, `protocols.3`) - a mover that given a receptor peptide pose mutates the peptides edge residues to cysteines, if needed, and enforces disulfide bonding by constrained minimization of the bond and the interaction. (again, thanks to @orlypolo )
- `FilterReporterMover` (`protocols/moves`, `protocols.1`) - a `FilterMover` that also calls `report()` on `apply()`
- `PeptideDeriverOutputter` (in `PeptideDeriverFilter.hh`) - a abstract base class for handling calculation outputs from `PeptideDeriverFilter`. Since `PeptideDeriverFilter` might have a set of outputs for each residue, for each chain in each chain-pair, outputting is quite elaborate. I figured I should decouple the calculation from the representation of results. Implementors of this class are responsible to implement representation of results. Subclasses are: (a) `PeptideDeriverMarkdownOutputter`, (b) `PeptideDeriverBasicOutputter` - output data from the protocol to a stream, in either a human-readable (markdown) or easily parsable (basic) format. (c) `PeptideDeriverPoseOutputter` - outputs important poses (using JD2 if possible). (d) `PeptideDeriverOutputterContainer` - delegates calls to a list of outputters; allows aggregation to allow multiple outputs from one `PeptideDeriverFilter` instance.
notify author
notify list [rosetta-logs@googlegroups.com]