The code allows for the generation of a molecular representation for amines used in carbon capture, generation from molecular fragment combinations and analysis of the chemical space. Documentation of the functionality can be found https://jammyzx1.github.io/Carbon-capture-fingerprint-generation/
Copyright IBM Corporation 2022. SPDX-License-Identifier: MIT
This toolkit is a collection of tools designed to assist in the design and prediction of carbon capture molecular properties. The current capabilities are:
- Chemical space plotting and analysis
- Chemical fingerprint generation
- Molecular dataset analysis
- Duplication of molecular string data identification
Please cite if you use the code:
@article{mcdonagh2024chemical,
title={Chemical space analysis and property prediction for carbon capture solvent molecules},
author={McDonagh, James L and Zavitsanou, Stamatia and Harrison, Alexander and Zubarev, Dimitry and van Kessel, Theordore and Wunsch, Benjamin H and Cipcigan, Flaviu},
journal={Digital Discovery},
volume={3},
number={3},
pages={528--543},
year={2024},
publisher={Royal Society of Chemistry}
url={https://pubs.rsc.org/en/content/articlehtml/2024/dd/d3dd00073g}
}
Please make contributions to the dev branch and open PRs to merge in to master branch.
We use docstring and unit tests to help maintain the library these are called through the unit_test.py script.
Please make sure all tests pass and add new tests for new code.
Once you have installed Anaconda, run the following commands
git clone $THIS_REPO
# careful, removes previous environment with the same name
yes | conda create --name ccsfp python=3.8
conda activate ccsfp
python setup.py install
python unit_test.py
The notebooks directory has examples which can be run to check the code runs correctly.