Proteuspy

Added

- Additional work on rcsb_viewer.py
- Automation scripts for Docker builds

Added

- One can now access a disulfide by name directly from the loader with:


pdb = Load_PDB_SS()
ss = pdb["2q7q_75D_140D"]

In prior versions one would need to use the loader.get_by_name() function.

Removed

- Removed the ``programs/rcsb_viewer.py`` program. The viewer now lives only in the ``viewer`` directory and can be invoked directly from the command line with:

console
$panel serve ~/repos/proteusPy/viewer/rcsb_viewer.py --show

Added

- Unified the disulfide viewers such that the rcsb_viewer.py program will work either stand-alone or in Docker.
- Added workflows to build the Docker images on GitHub and Docker Hub
- Made pyqt5 an optional install, pip install proteusPy[pyqt5] adds it back.

Added

- Renderers:
- Docker image for the pyVista renderer. This lives under proteusPy/viewer and is deployed on Docker hub under ``egsuchanek/rcsb_viewer``.
Launch with: ``docker -d -p 5006:5006 egsuchanek/rcsb_viewer:latest. Works under MacOSX and Linux.
- The standalone Panel-based version lives in ``proteusPy/programs/DBViewer.py``. Launch with: ``panel serve path_to_DBViewer.py --autoreload &``
- PyQt5 version. This lives in ``proteusPy/programs/QT5Viewer.py``. It is the most advanced version, but I'm unable to build under Linux. My intent was to deploy this via Docker, but can't get PyQt5 to build currently.

- ``DisulfideList.center_of_mass`` - returns the ``Vector3D`` center of mass for the list.
- ``DisulfideList.translate()`` - adds the input ``Vector3D`` object from the list, destructively modifying the coordinates. Used primarily in rendering functions to center the list at ``Vector3D([0,0,0])``.
- ``Disulfide.translate()`` - translates the input Disulfide by the input ``Vector3D``.

Issues

- I cannot get the QT5 viewer to build under linux. The pyQt5 library won't install.

Added

- Disulfide QT5 viewer development, improvement
- ``DisulfideList.display()`` added to provide a summary of the input DisulfideList
- Additional analytics


Fixed

- Analysis of the PDB entry files revealed yet another parsing issue. There are many structures that contain disulfides referring to themselves, ie 25A-25A. These had not been caught in any prior release and were revealed while I was working on the filtering code.

Added

Created a Disulfide viewer using pyQt5 library. This program is under programs/viewer.py. It's basic, but provides an easy way to visualize disulfides within the database. The single checkbox toggles between individual disulfide, or overlaid. The latter shows all of the disulfides for the given protein entry, overlaid onto a common coordinate system. I continue to tweak the code, but the core is stable.