Merge pull request #4018 from RosettaCommons/JWLabonte/PTMs/database
Quick bug fix: removing CRLFs that snuck in
This merge is a simple file-compatibility bug fix.
It cannot possibly affect any tests.
notify author
notify list [rosetta-logs@googlegroups.com]
Merge pull request #4004 from RosettaCommons/vmullig/threadsafe_options_system
Make the options system threadsafe
The options system is a nasty hunk of nonconst global data that isn't threadsafe. Let's make it threadsafe.
Tasks:
- [x] Make `OptionCollection` class threadsafe.
- [x] Including the global variables defined in `OptionCollection.cc`.
- [x] Make `ScalarOption_T_` threadsafe.
- [x] Make `VectorOption_T_` threadsafe.
- [x] Make `Option` class threadsafe.
- [x] Add a `utility::thread::PairedReadLockWriteLockGuard` class for simultaneously read-locking one mutex and write-locking another without deadlock (_e.g._ for copying values from one object to another).
- [x] Make the mutex-locked versions of the functions in `VectorOption_T_` private.
- [x] Beauty.
Possible additional spots to make threadsafe in a future pull request:
- Make `AnyOption` class threadsafe.
- Make `AnyOptionVector` class threadsafe.
- Make `utility::keys::AutoKey` class threadsafe.
notify author
notify list [rosetta-logs@googlegroups.com]
Merge pull request #4003 from RosettaCommons/jkleman/SciBen_fixes
Jkleman/sci ben fixes
Most tests pass: https://benchmark.graylab.jhu.edu/revision?branch=&id=12668
* except antibodies - Jeli is looking into it
* mp_dock and mp_symdock - debug passes (https://benchmark.graylab.jhu.edu/revision?branch=&id=12684), release is still failing, requires Rebecca's expertise (asking her to fix those when merging her new ones in)
* PDB diagnostic - in Steven's hands
* docking wasn't run previously
* ligand docking - is in Shannon's hands
* stepwise RNA fixed for debug mode: https://benchmark.graylab.jhu.edu/revision?branch=&id=12686
notify author
notify list [rosetta-logs@googlegroups.com]
Merge pull request #3994 from RosettaCommons/JWLabonte/PTMs/parsing
EnzymaticMovers: Expanding parsing of AA sequences
This merge will expand the parsing capabilities of `EnzymaticMover`s by adding additional one-letter codes (B, J, O, U, Z) for amino acid residue pairs or NCAAs and allowing `X[]` to specify a specific `ResidueType` by 3-letter code.
- The parser will now recognize the IUPAC-approved one-letter codes B, J, O, U, and Z, which code for Asx, Xle, Pyl, Sec, and Glx, respectively.
- X alone is recognized to be any of the 20 canonical amino acids; X followed by square brackets specifies a single non-canonical amino acid by 3-letter code. For example, `X[SEP]` specifies phosphoserine.
- Parentheses are used to specify multiple possible residue types at that site, separated by forward slashes, _e.g._, `(A/G)` specifies either Ala or Gly at that position.
notify author
notify list [rosetta-logs@googlegroups.com]
Merge pull request #3999 from RosettaCommons/JWLabonte/PTMs/cofactors
Cofactors: turning off loading of PDB components for ATP
This merge will cause Rosetta to use the ATP in its own database, which doesn't have protonated phosphate oxygens, instead of the one from the PDB components library, which does.
notify author
notify list [rosetta-logs@googlegroups.com]
Merge pull request #4001 from RosettaCommons/jaumebonet/coupledmovesRS
CoupledMovesProtocol: fixed attribute definitions between parse_my_tag and provide_xml_schema
notify author
notify list [rosetta-logs@googlegroups.com]
Merge pull request #3985 from RosettaCommons/JackMaguire/AddFirstTFTest
Adding a tensorflow test that loads in a very simple neural network and evaluates it. The test fails if the output of the network is not what we expect
notify author
notify list [rosetta-logs@googlegroups.com]