scsdpy 0.1.0

SCSD is Python software for the analysis of molecular conformation and deformation in crystal structures

SCSD

Symmetry-Coordinate Structural Decomposition - a tool for crystal structure analysis Dr. Christopher J. Kingsbury and Prof. Dr. Mathias O. Senge

Introduction

The scsd (symmetry-coordinate structural decomposition) program is a method of describing the concerted distortions of molecules in crystal structures using the idealised point group symmetry. This can be used for a number of purposes - including structure description, comparison and generation.

The latest version can be downloaded using the following command:

pip install scsdpy

This program can be accessed through a notebook interface by importing the underlying module. Many users may find operating through the web interface at https://www.kingsbury.id.au/scsd accessible and suitable. The analysis of known molecules can also be performed with scsd_mercury.py through the Mercury interface, with more information available on the ccdc-opensource repository (https://github.com/ccdc-opensource)

This work can be cited as below:

Kingsbury CJ, Senge MO. Quantifying Near-Symmetric Molecular Distortion Using Symmetry-Coordinate Structural Decomposition. ChemRxiv. 2024; doi:10.26434/chemrxiv-2024-6b25s

This citation will be updated after peer review.

While this repository contains tables referring to point groups other than those listed (C_2v,C_2h,D_2d,D_2h,D_3h,D_4h,D_6h) these are not verified; check that the atom positions and their sums are in alignment and that the numbers make sense. Use at own risk.

Examples

A tutorial notebook is included to demonstrate model generation and analysis, database investigation and conformer generation.

• notebooks/
  • scsd_tutorial.ipynb
  • data
    • bevron.pdb
    • puybaf.pdb
    • pispco.pdb
    • robsons.sd

Examples of integration of these tools with the csd-python-api interface are demonstrated; scsd_mercury.py and scsd_model_mercury.py are intended to be run through the csd-python-api menu in CCDC-Mercury.

scsd_collection_mercury.py and scsd_single_cl.py contain methods to import data through this interface, and requires a ccdc license.

• scsd_mercury.py
• scsd_model_mercury.py
• scsd_collection_mercury.py

Dependencies:

• numpy
• scipy
• seaborn
• plotly
• pandas
• networkx
• flask

Optional

• ipympl
• scikit-learn
• werkzeug
• ccdc

This work is licensed under THE ANTI-CAPITALIST SOFTWARE LICENSE (v 1.4)

To view a copy of this license, visit https://directory.fsf.org/wiki/License:ANTI-1.4 Commercial or military use is forbidden.

Please contact Chris Kingsbury ([e-mail]) for collaboration or with any questions

Ongoing issues

To Do List:

• make 3-fold axis PCAs make sense - there isn't uniformity, and no way to 'normalise' currently
• comments on the main program aspects
• "Lacunary" - how to model when a part is missing? FeMoCo is part of this
• extend sensibly to O_h ... ongoing work