This is an experimental port of the APBS Tools plugin for PyMol version 2.x. It's under development and not ready for production use; I'm writing it as an exercise to learn PyQt.

Note that this is only relevant for the open-source fork of PyMOL; this functionality is included by default in the incentive fork of PyMOL provided by Schrodinger.

The APBS (Adaptive Poisson-Boltzmann Solver) Tools plugin is used to compute macromolecular electrostatic charges and visualize this information as a potential surface in PyMOL. It does this by calling the apbs code; optionally, pdb2pqr is called to preprocess input files; see poissonboltzmann.org.

The current version of the plugin (version 2.1) was written against the tkinter GUI library used in PyMOL 1.x; version 2.x of PyMOL migrated to PyQt. There is legacy support for tkinter-based plugins (not implemented in the APBS Tools plugin), but I wanted to take this as an opportunity to update it to PyQt.

Since development is still in progress, currently the only supported installation method is through conda. If conda isn't installed, download Miniconda3.

If you're installing on an arm-based mac, you'll need

export CONDA_SUBDIR="osx-64"

since arm-based builds of apbs aren't available on conda (and my attempts to build from source have failed so far.) Note that conda packages for apbs have fallen behind tagged releases.

Then do

conda env create -f=conda.yml; conda activate APBS_Qt_Plugin

To add the plugin in PyMol, select the APBS_Qt_plugin directory in the Plugins manager.

Original UI	UI in incentive fork	This code

I've attempted to reorganize the code along the lines of the (hierarchical) Model-View-Controller pattern. This is absolutely overkill for an application whose scope is this small, but I wanted to use the project as a way to explore coding practices that scale to large applications.

At least subjectively, this has felt like "going against the grain" of PyQt's intentions. Qt's use of Model/View terminology refers to something subtantively different, as clarified in this SO post. UI widgets operate according to an event-driven paradigm using signals and slots, but neither PyQt's *Model objects or the base data types wrapped from C++ Qt have predefined signals or slots: these need to be defined manually on the Model class (as in this example), which results in boilerplate code and redundant definitions.

The solution I've used here (perhaps the most novel part of this project) is to combine a solution proposed here with the attrs package to automate defintion of Signals and Slots on the Model classes. This is done in util.py.

The original APBS Tools plugin was written by Michael G. Lerner in 2009, with contributions from Heather A. Carlson and Warren L. DeLano. The current version is available here.

Users of apbs and pdb2pqr are strongly encouraged to register their use at poissonboltzmann.org, as this is key to securing funding for continued development of this software.

Citation for apbs:  Baker NA, Sept D, Joseph S, Holst MJ, McCammon JA. Electrostatics of nanosystems: application to microtubules and the ribosome. Proc. Natl. Acad. Sci. USA 98, 10037-10041 (2001); doi/10.1073/pnas.181342398; PMID: 11517324; PMCID: PMC56910.

Citation for pdb2pqr:  Dolinsky TJ, Nielsen JE, McCammon JA, Baker NA. PDB2PQR: an automated pipeline for the setup, execution, and analysis of Poisson-Boltzmann electrostatics calculations. Nucleic Acids Research 32 W665-W667 (2004). doi/10.1093/nar/gkh381; PMID: 15215472; PMCID: PMC441519.