Merge pull request #2979 from RosettaCommons/kalekundert/core_rosetta_scripts_utils
Move parse_task_operations(), parse_score_function(), etc. into core.
notify author
notify list [rosetta-logs@googlegroups.com]
Merge pull request #2975 from RosettaCommons/vmullig/fa_dun_ncaa_mirror_scoring_hotfix
Fixing a *very* subtle bug introducing minor asymmetry in NCAA fa_dun scoring
When I wrote the Voronoi-based well assignment code for the NCAA fa_dun scoring, I introduced a small bug: for mirror-image types (_e.g._ D-amino acids), I was forgetting to flip the mainchain torsion values when figuring out which backbone bin I was in. This meant that the boundaries between rotamer wells could be very slightly out of place. I discovered this in the peptoids branch, since it becomes most apparent in types in which the rotamer well centres change most with backbone torsion value changes.
This pull request fixes the problem.
notify author
notify list [rosetta-logs@googlegroups.com]
Merge pull request #2966 from RosettaCommons/JackMaguire/MCHBNetCorrection
This PR introduces an option to HBNet that allows HBNet's internal scores to be placed in the output poses (turned off by default). This would allow the user to access the scores using the CalculatorFilter or ReadPoseExtraScoreFilter and the scores would be included in the score.sc file.
notify author
notify list [rosetta-logs@googlegroups.com]
Merge pull request #2973 from RosettaCommons/sergey/binder
Updating Binder submodule, adding include mapping for GCC regex headers. Switching PyRosetta build script to use SSH for Pybind11 and LLVM checkout. Triggering clean build.
notify author
notify list [rosetta-logs@googlegroups.com]
Merge pull request #2969 from RosettaCommons/raemisch/BugFixSimpleGlycosylateMover
SimpleGlycosylate copy constructor changed to include proper residue selector.
notify author
notify list [rosetta-logs@googlegroups.com]
Merge pull request #2972 from RosettaCommons/JackMaguire/MultistageRosettaScripts
There was a bug in an integration test I added to master yesterday that causes it to always fail. I did not notice it because the testing server did not have a reference to "diff" against, but now that other people are pushing to master I can see that all the diffs are backtraces. My apologies to everyone who had to deal with these useless test failures!
notify author
notify list [rosetta-logs@googlegroups.com]
Merge pull request #2959 from RosettaCommons/JackMaguire/MultistageRosettaScripts
Adding Clustering Feature in Multistage Rosetta Scripts
I am adding the ability for Multistage Rosetta Scripts to increase diversity by clustering results between stages. I will talk about this more at winter rosettacon next week if you are interested.
I expect it to look something like this:
```
<PROTOCOLS>
<Stage num_results_to_keep="2000" >
...
</Stage>
<Cluster num_results_to_keep="1000">
<SomeKindOfMetric/>
</Cluster >
<Stage>
...
</Stage>
</PROTOCOLS>
```
notify author
notify list [rosetta-logs@googlegroups.com]
Merge pull request #2946 from RosettaCommons/vmullig/hbnet_energy_options
Add alternative ramping options for the HBNet energy
The HBNet energy currently gives a quadratically-ramping bonus for larger and larger hydrogen bond networks. @basantab observed that this results in giant hydrogen bond networks if there is any way at all to form these, at the expense of bad packing or other undesirable features. A "diminishing returns" ramping function might be better.
This pull request adds options for linear, logarithmic, and square root ramping. It also adds an option for a maximum network size, beyond which there's no bonus.
Tasks:
- [x] Add enum for the new ramping types.
- [x] Add different calculators for the bonus function.
- [x] Add means for setting ramping type from flags.
- [x] Add means for setting ramping type from RosettaScripts.
- [x] Unit tests for different ramping types.
- [x] Add option for max network size.
- [x] Modify calculator for max network size.
- [x] Add means for setting max network size from flags.
- [x] Add means for setting max network size from RosettaScripts.
- [x] Unit tests for max network size.
- [x] Documentation for ramping types.
- [x] Documentation for max network size.
- [x] Beauty.
notify author
notify list [rosetta-logs@googlegroups.com]
Merge pull request #2947 from RosettaCommons/sboyken/keep_info_labels_symmetric
options to preserve PDBInfoLabels when detecting / generating symmetric poses; works with SetupForSymmetry and DetectSymmetry movers by simply adding keep_pdb_info_labels="true" to these movers. Longterm would be good to decide whether people want this to be default, and could be hooked into deeper symmetry code
notify author
notify list [rosetta-logs@googlegroups.com]