[update of SandwichFeatures] - more clear naming (primary_seq_distance_cutoff_for_electrostatic_interactions -> min_primary_seq_distance_diff_for_electrostatic_interactions) - more consistent naming (shortest_dis_between_facing_aro -> shortest_dis_between_facing_aro_in_sw) - more consistent naming (shortest_dis_between_sheets_by_aro -> shortest_dis_between_facing_aro_in_sw) - may allow to exclude sandwich that has too close distance between facing aromatic residues - like 'write_resfile_to_minimize_too_many_core_heading_FWY_on_edge_strands_' no aro for every 2nd residue for 'write_resfile_to_minimize_too_many_core_heading_FWY_on_core_strands_' - should not change any integration tests other than (expected) features and features_database_schema (if any)

• Summary

Remove unused sugars/legacy file category in src.settings file to fix Xcode project.

• Summary

Trying to fix debug-mode MPI build (gcc/Linux). Unit test status: Not Run Integration test status: Not Run I didn't run tests because these are trivial changes (all commenting out unused variables).

• Summary

Minor change to claim strength of FragmentJumpClaimer. Nearly trivial change in untested code; no test changes expected.

• Summary

Trying to fix signed/unsigned comparison warnings-as-errors in the MPI build.

• Summary

Automatic metal setup and constraints now works within RosettaScripts. CAVEAT: Constraint weights must be turned on explicitly in scorefunctions defined within a RosettaScript. Unit test status: All pass. Integration test status: Not run.

• Summary

ResidueType: added VIRTUAL_RESIDUE property & fixed is_virtual_residue() VRT and VRT1 now have the VIRTUAL_RESIDUE property. ResidueType now has a is_virtual_residue_ private datum. is_virtual_residue() now returns the value of the latter. All unit tests pass. No integration test changes.

• Summary

small fix for -in::missing_density_to_jump

• Summary

restoring some units tests broken by change in patch linker from "_p:" to ":". A few more tests remain broken; will commit fixes later in the day.

• Summary

Updating Windows/BOINC project files and update scripts to VisualStudio 12.0 format (VS2013).

• Summary

Fixing Windows build (VS2013).

• Summary

updates to pilot apps. fix debug build

• Summary

the HA atoms in centroid mode did not (anymore) reflect the full-atom definitions. This causes non-ideal poses when switching to full-atom from centroid with HA. I suppose in the mean-time the full-atom definitions of ideal residues have diverged. Now the HA patch is exactly the same as fa definitions (again).

• Summary

Big merge -- a few little improvements for entire code, and several important updates specifically for stepwise assembly and monte carlo code. All integration tests checked very carefully. In terms of affecting lots of apps: (1) Patch linker text is now ':', not '_p:', which was more verbose and (I found) confusing; (2) Number of default residue types has been reduced a lot, by putting patches under '-patch_selector' flags, and a warning message is triggered if a dev starts a patch explosion; (3) got rid of silly warning message from hbonds about DonorStrength.csv missing. Also, fixed an initialized variable in LigandConformation class caught by instability in cstfile_to_theozyme_pdb test (and by valgrind). In SWA/SWM, almost everything is seamlessly unified between RNA & protein modeling so that RNA/protein (and more general cases, including ligands & metal ions) can be easily set up within the stepwise framework, finally.

• Summary

Prevent ConstraintFragmentSampler to always set cst_weight to default value if not specified by user. Cst_weight is now only set, if indicated by user (and thus not overwriting atom_pair_constraint patches). Expected Integration Test Changes (Overall Score is identical, only the atom_pair_constraints=0 entries are removed): broker ThreadingInputter broker_membrane fold_and_dock metalloprotein_broker

• Summary

fix a bug when metropolis_hastings runned with FixTemperatureController

• Summary

Legend: