Merge pull request #2774 from RosettaCommons/sergey/f2
Adding emplace constructor to NamedAtomID
Vikram K. Mulligan 7 years The template line added to NamedAtomID.hh in this commit seems to be tripping up the beautifier. Andrew -- can you take a look at this?[list]
notify author
notify list [rosetta-logs@googlegroups.com]
Merge pull request #2775 from RosettaCommons/vmullig/indexselector_noerror
Add an option to the ResidueIndexSelector to allow it to ignore residues that aren't in the pose
Currently, the `ResidueIndexSelector` throws an error if a residue index specified does not exist in the pose (_e.g._ residue 56 of a 55-residue pose). While this is a good default behaviour, users have encountered situations in which this is not a desired behaviour. This pull request retains the current behaviour as default behaviour, but adds an option to disable this.
Tasks:
- [x] Add option to ResidueIndexSelector.
- [x] Setters & getters.
- [x] Update `parse_my_tag()`.
- [x] Update xsd functions.
- [x] Update unit test.
- [x] Update serialization functions.
- [x] Beauty.
- [x] Documentation.
@srgerb
notify author
notify list [rosetta-logs@googlegroups.com]
Merge pull request #2769 from RosettaCommons/roccomoretti/icc_scons_fix
Fix compiler autodetection for ICC.
Add os information to pre-load stage, such that the OS-specific ICC commands can be loaded and used for pre-load detection.
notify author
notify list [rosetta-logs@googlegroups.com]
Merge pull request #2766 from RosettaCommons/sergey/binder
- Adding dependency tracking for both generation-phase (Binder invocation) and CMake invocation.
- Moving CMake invocation into generate_cmake_file so build.py -s immediately call ninja instead of always re-running CMake first.
- Using symlinking instead of copying for PyRosetta Python files.
After this is merged i expect that building PyRosetta with just build.py or optionally build.py -s (to skip generation phase entirely to win extra ~1.5s) should became the best option for most developers.
notify author
notify list [rosetta-logs@googlegroups.com]
Merge pull request #2754 from RosettaCommons/jadolfbr/glycan_scanner3
Design Tools
A number of design tools are added in this PR, including some specifics for glycans.
New Resfile Commands
=================
First, we have 3 new resfile commands: `CHARGED`, `AROMATIC`, and `PROPERTY`.
The third is a general command that takes any ResidueProperty. Currently, it only works for Cannonicals, but perhaps that could be generalized int the future for NCs.
New General TaskOps/Movers
======================
- `ResfileCommandOperation` : This operation takes a residue selector and ANY resfile command operation line (without residue numbers) and applies it. It is incredibly powerful. For example, You can select all beginning of helices and design them as charged residues extremely easily. This works with NCs as well.
```
USAGE:
<ResfileCommandOperation name=(string) command=(string) residue_selector=(string)>
</ResfileCommandOperation>
```
- `SequenceMotifTaskOperation` : This is an operation that applies a given _design sequence motif_. This motif is kinda-sorta like regex syntax, but a bit different. It accepts a residue selector to design in this sequence motif starting at each position specified in the selector. Also takes resfile commands. The syntax for the operation is as follows:
```
<SequenceMotifTaskOperation name=(string) motif=(string) residue_selector=(string)>
</SequenceMotifTaskOperation>
EXAMPLE:
Glycosylation N-Linked glycosylation motif design: N[^P][ST]
motif (string): This is slightly similar to a regex, but not quite.
We are not matching a sequence,
we are designing in a motif regardless of the current sequence,
anywhere in a protein.
- Each letter corresponds to a position. Using [ ] indicates a more complicated expression
for that position.
- An X indicates it can be anything, and that we are designing here.
- An AA Letter, like V, indicates that that position will be designed to a V.
- A - charactor indicates that that position stays with whatever it is currently.
We essentially skip this position.
- An expression like: [^PAV] indicates that we will design anything except
Proline, Alanine, and Valine
- An expression like: [NTS] indicates that that position can be Asparigine,
Threonine, or Serine and
only of these will be enabled during the design.
- RESFILE commands are accepted as well.
These require a % charactor in from of the whole expression.
For example [%POLAR] would set that position to only polar design.
This is exactly the same as a resfile line, so you can even do NC like so:
[%EMPTY NC R2 NC T6 NC OP5]
```
- `CreateSequenceMotifMover` : Applies the operation, using the packer, packing neighbors by default. Has a few other options for common uses.
Glycans
======
- `CreateGlycanSequonMover` : This calls the SequenceMotifMover with pre-specified motifs, including using the enhanced motif that seems to result in higher glycosylation and less complex glycans ( references included in descriptions).
Other
========
New string_util functions
Updates to code templates
Integration tests for movers
Unit tests for operations, and each individual function including the string utility functions.
notify author
notify list [rosetta-logs@googlegroups.com]