Merge pull request #2657 from RosettaCommons/bcov/NeighbohoodUpdateGraph
NeighborhoodResidueSelector clone/update_residue_neighbors() + some util functions
Sergey Lyskov 7 years Brian, looks like your commit changed the head of Binder submodule (source/src/python/PyRosetta/binder). And this leads to PyRosetta builds failure. Could you please looks this up?
notify author
notify list [rosetta-logs@googlegroups.com]
Merge pull request #2661 from RosettaCommons/aleaverfay/add_no_sc_to_ia_option
Avoid shape complementarity calculation due to crash in InterfaceAnalyzer
Shape-complementarity score is reliably crashing on my cluster, but works fine
on my workstation, so I am adding a short circuit in the interface analyzer:
put "-compute_interface_sc false" on your command line to skip that calculation.
I'm trying to debug the problem, but valgrind is turning up jack squat. In the mean time, I need InterfaceAnalyzer to work.
Vikram K. Mulligan 7 years Note that there are known issues with shape complementarity and metals (or other rarer atom types). As I recall, the shape complementarity filter implements its own definition of atom types, rather than using the Rosetta ones. Could this be related to the failure?
notify author
notify list [rosetta-logs@googlegroups.com]
Merge pull request #2398 from RosettaCommons/vmullig/energy_based_clustering
Make "bettercluster" (now "energy_based_clustering") into a public app in anticipation of publication
Tasks:
- [x] Move to public apps.
- [x] Rename to "energy_based_clustering" (internally, too).
- [x] Add copyright info.
- [x] Correct tracer names.
- [x] Move bulk of functionality to protocols.
- [x] Trim app down to a bare-bones thing that calls functionality in protocols.
- [x] Move options to `options_rosetta.py`.
- [x] Add to options_rosetta.py.
- [x] Remove from app.
- [x] Update app to take an options object, with option to set from global options system.
- [x] Disable PCA analysis for now.
- [x] Check generality -- there may be old hard-coded assumptions about mainchain torsion names, _etc_.
- [x] Unit tests.
- [x] Cyclic alpha/oligourea peptide, Cartesian-based clustering.
- [x] Cyclic alpha/oligourea peptide, dihedral-based clustering.
- [x] Cyclic alpha/oligourea peptide, with symmetry, Cartesian-based clustering.
- [x] Cyclic alpha peptide, Cartesian-based clustering with cyclic permutations.
- [x] Cyclic alpha peptide, dihedral-based clustering.
- [x] Integration test -- cyclic peptides.
- [x] Integration test -- torsion-based clustering.
- [ ] Documentation.
- [x] Beauty.
Note: the `vmullig/oligourea` branch has been merged into this branch. As such, pull request #2528 must be merged before this one -- DONE.
For the future:
- Check that this works for Jacob's test cases (fragments of proteins).
- Check that this works for Christine's test cases (linker-containing peptides).
- Check that this works for Parisa's test case (which caused a segfault before).
- Check that this works for Tim's test cases (N-methyl amino acids).
notify author
notify list [rosetta-logs@googlegroups.com]
Merge pull request #2655 from RosettaCommons/vmullig/fix_oligourea_design_test
Fix oligourea_design integration test
The SCORE lines in the output silent file currently contain timing information. They ought not to. Andy pointed me to the -run:no_prof_info_in_silentout flag to disable this. (Thanks, Andy!)
notify author
notify list [rosetta-logs@googlegroups.com]
Merge pull request #2640 from RosettaCommons/roccomoretti/JobOutputterFactoryFix
Fix issue with missing mmCIFJobOutputter
For score_jd2, the mmCIFJobOutputter case wasn't being called, due to copy-paste coding.
Centralize the JobOutputter logic, to coordinate these two cases.
notify author
notify list [rosetta-logs@googlegroups.com]
Merge pull request #2563 from RosettaCommons/roccomoretti/compound_filter_details
Permit CompoundFilter to give more info about which subfilters are failing/succeeding.
To do this with minimal user-facing perturbation, I had to reorganize how the confidence value was treated in FilterFactory (basically, remove the usage of CompoundFilter there).
This also involved overhauling the StochasticFilter -- the semantics of StochasticFilter were messed up as a side effect of how the confidence value was being treated. I regularize/rationalized the StochasticFilter interface, then added the ability to run a subfilter on StochasticFilter success/failure, and re-did the FilterFactory to use that directly, rather than the StochasticFilter/CompoundFilter approach it used before.
Unit tests for StochasticFilter and FilterFactory were added, to make sure I got things right. Some slight trajectory changes in integration tests, as this change reduces the number of random number draws in the `confidence="0"` case.
notify author
notify list [rosetta-logs@googlegroups.com]
Merge pull request #2623 from RosettaCommons/BYachnin/set_catalytic_respack_behavior_xsd
Added XSD info for SetCatalyticResPackBehavior
Failed tests are almost surely due to other commits to master that have changed several integration tests.
notify author
notify list [rosetta-logs@googlegroups.com]