Fixing Mac PyRosetta build

• Summary

Merge pull request #110 from RosettaCommons/PyRosetta Fixing PyRosetta build on Linux

• Summary

Merge pull request #109 from RosettaCommons/JWLabonte/MinMover MinMover fix for score_function() setter for PyRosetta No unit test changes. Integration tests show the usual suspects of broken tests and trajectory changes in KIC-using tests but only in release_debug.

• Summary

Fix cmake build w/r/t recent pointer changes.

• Summary

Merge pull request #104 from RosettaCommons/jadolfbr/antibody_design Updates to antibody and antibody design. Merge Pull Request #104 This branch mainly has updates to the antibody_designer app and RosettaAntibody. It has been going for quite a while, so here goes. This is a complete refactor of antibody/design with a few major additions to the app (controlled either via cmd-line options or the Instructions File): * optional Paratope/Epitope site-constraints with cdr and rsd-lvl control respectively * optional framework and antigen design now integrated. * neighbor CDR minimization during the GraftDesign stage (via instructions file) * SeqDesign during GraftDesign (via instructions file) * ability to limit CDRs used for GraftDesign stage by species and germline (via instructions file) * Lambda/Kappa light-chain option set through cmd-line, held in AntibodyInfo, accessed by AntibodyDatabaseManager to limit CDRs * lots of small optimizations, bug fixes, and options. =============== Antibody Database: * Added an SQLITE3 antibody database for general use to database/sampling/antibodies that uses data from our 2011 North CDR clustering paper (This is the default version for the app, if a newer one is not being used). * Newer, PDB-current versions of this database will be available here soon: http://dunbrack2.fccc.edu/PyIgClassify/ * Antibody Database classes/functions refactored and separated from design. Moved to antibody/database for general use. ========== AntibodyInfo: * added a few functions/refactoring that I was working on with Brian Weitzner in his antibody_refactor branch concerning loops and movemaps. * added AntibodyRegionEnum and function in AntibodyInfo for accessing the region of a particular residue (framework_region, cdr_region, antigen_region). This makes control of design regions much simpler to setup. =============== Code Refactor: * options now reside in their own classes at the CDR-level (CDRSetOptions, CDRSeqDesignOptions, CDRGraftDesignOptions), with their own Parsers for parsing instuctions files * AntibodyDesignMoverGenerator which aims to create/setup movers for antibody design, as well as run them via apply. * This will be expanded for use in RS, and foldtree stuff will eventually be replaced with the Broker system thanks to helpful chats with Justin Porter * Deprecates AntibodyDesignModeler * AntibodySeqDesTFCreator which sets up TaskOps and TaskFactories for use in different parts of the antibody design protocol from the set of CDRSeqDesignOptions. * new antibody/constraints namespace ===== Other: * Added/refactored docking setup_foldtree functions to have functions that don't automatically add the FoldTree to the pose. * Needed to move InterfaceInfo setup out of the setup_foldtree functions and into their proper place in DockingLowRes * Added option for neighbor design to RestrictToLoops/AndNeighbors * Added option to RestrictToLoops which basically makes it act as a RestrictDesignToLoops task op and does not touch packing. Simplified apply logic. * Added options to FastRelax to enable cartesian min and dualspace settings via code and via parse_my_tag. * Added debug option to SnugDock which switches the random mover to a sequencemover to aide in identification of bugs * Suite of helpful antibody design functions in antibody/design/util ================= Added/Modified Tests: Unit tests have been added for antibody design options classes, parsers, constraints, and utility functions. Unit tests for using the antibody database are included as well. Updates and additions to AntibodyInfo and RestrictToLoops unit tests. =================== Integration Test Changes: Integration test change in antibody designer (expected). Antibody integration tests and AnchoredPDBCreator show cosmetic changes. 3 tests are failing in release.debug as before this merge (app_exception, non ideal_rtmin, make_exemplar). Some tests (KIC_with_fragments, NGK, enzdes) show small numeric noise on the order of .01 and .001 A coord changes. ============== All Unit tests pass.

• Summary

Cleaning up full shared_ptr definitions in .cc files for better code readability. These were introduced by the automatic rewriting. Cosmetic syntax changes only. Compiles OK. No functionality or test changes expected.

• Summary

Updating master (local) with remote. This merges a fix for the membrane_ddG test (turned off) and also turns off the lipid accessibility file reading unit test (which will also be fixed in a commit soon to come. Nothing in the build should change here.

• Summary

Merge branch 'master' of github.com:RosettaCommons/main

• Summary

Fixing a handful of performance bugs following the utility::owning_ptr --> boost::shared_ptr conversion. Also inlining access to some data in ResidueType and AtomType that is needed in the innermost loops of atom pair energy evaluation. The 1000x_fa_atr performance benchmark shows a hefty improvement (2x).

• Summary

Fixing compile errors in crystal refine code

• Summary

Fixing cxx11thread build. Changing settings to build with cxx11 std:: pointers by default. Requires c++11 capable compiler to build with extras=cxx11thread; tested with gcc 4.8. Couple fixes to the MultiThreadedJobDistributor which didn't get automatically rewritten.

• Summary

Fixing mpi build.

• Summary

Fixing PyRosetta build settings (hopefully).

• Summary

AST-based rewrite of 95b9f06c8a82ad9bfca4d6e9ff19ce04bfb631f5 for CXX11, plus some minor build fixes. This commit now compiles with boost (default) and cxx11 STL pointers (extras=cxx11). ==> Builds: clang debug: OK clang debug cxx11: OK clang release: OK clang release cxx11: OK gcc debug: OK gcc debug cxx11: OK gcc release: OK gcc release cxx11: OK pyrosetta: not tested mpi: not tested static: not tested mysql/postgres: not tested ==> Tests with clang 3.5: Unit tests: All pass Integration tests (vs. original master in 57300): - Expected IG size changes (8 byte decrease on 64-bit) - Trajectory changes: swa_rna_gagu_04_clustering - Cosmetic tracer output changes as in test 57301 - All tests run without crashes ==> Tests with clang 3.5, cxx11: Unit tests: All pass Integration tests (vs. original master in 57300): - Expected IG size changes (8 byte decrease on 64-bit) - Many trajectory changes (not unexpected), see PR #82 for additional notes - All tests run without crashes ==> Tests with gcc 4.8.2 and gcc 4.8.2, cxx11: Unit tests: All pass Integration tests (vs. original master in 57300): - Expected IG size changes (8 byte decrease on 64-bit) - Many small trajectory changes - Fewer "core.optimization.Minimizer: WARNING: DFPMIN (Armijo) MAX CYCLES 2000 EXCEEDED, BUT FUNC NOT CONVERGED" messages in swa* - Cosmetic tracer output changes as in test 57301 - All tests run without crashes

• Summary

Fixing PyRosetta 'basic' unit test as requested by Luki

• Summary

Changes from branch luki-mbi/pointer_rewrite (PR #80). ==> Builds: clang debug: OK clang release: OK gcc debug: OK gcc release: OK pyrosetta: not tested mpi: not tested static: not tested ==> Tests with clang 3.5: Unit tests: All pass Integration tests: 2 cosmetic changes: fold_and_dock ThreadingInputter ==> Tests with gcc 4.8.2: Unit tests: All pass Integration tests: 69 changes Many trajectory changes, mostly to swa* Some cosmetic changes (as above)

• Summary

Legend: