Merge pull request #1472 from RosettaCommons/jadolfbr/jadolfbr/carbohydrates/val_fix
Fix simple glycosylation
Fix an uninitialized variable in the copy constructor of glycan relax that was causing the simple glycosylation unit test to fail in valgrind. Thanks to @roccomoretti for actually debugging this and finding where the variable was uninitialized in the first place.
notify author
notify list [rosetta-logs@googlegroups.com]
Merge pull request #1471 from RosettaCommons/dimaio/app_bugfixes
Fixing two bugs in hybridize/cartesian_sampler. Thanks @roccomoretti for finding these.
notify author
notify list [rosetta-logs@googlegroups.com]
Merge pull request #1467 from RosettaCommons/roccomoretti/valgrind_benchmark_fix
More integration.py fixes
Fix integration.valgrind -- slightly different flags are needed. (--skip-comparison messes it up.)
Also, fix the integration test execution time ordering. (The extra directory in the test name messed things up.)
notify author
notify list [rosetta-logs@googlegroups.com]
Merge pull request #1454 from RosettaCommons/roccomoretti/alignVectorSets
Fixes for numeric:alignVectorSets
numeric:alignVectorSets() had some edge conditions (zero vectors, co-linear vectors) which resulted in utility_exits. Add special cases to catch these conditions and do something reasonable instead.
Unit tests for alignVectorSets() added. There's some integration test changes observed (KIC_with_fragments, density_denovo), which look to be due to changes in optimizations w/r/t this inlined function. (The changes don't happen locally with a debug build.)
notify author
notify list [rosetta-logs@googlegroups.com]
Merge pull request #1465 from RosettaCommons/amelie-stein/ddG_with_ligands
On behalf of Amelie, merge in some fixes to the ddg_monomer applicaition to allow it to work with non-amino acid residues.
notify author
notify list [rosetta-logs@googlegroups.com]
Merge pull request #1458 from RosettaCommons/vmullig/ncaa_betanov
Fix loading of NCAA rotamer libraries when using beta_nov15
The problem turned out to be that beta_nov15 turns on the Shapovalov rotamer libraries, which have a slightly different file format (an extra column in the middle of the file, which is irritating). When an NCAA library is loaded, it is therefore misinterpreted: Rosetta assumes that the extra column is there when it is not. The solution is to count the columns in the first line and to determine whether a file is a Shapovalov-format file or not.
Tasks:
- Fix the bug.
- Merge the latest master into branch.
- Fix the fix so that it doesn't break the loading of gzipped rotamer libraries.
- Beauty.
notify author
notify list [rosetta-logs@googlegroups.com]