A graph representation of molecular ensembles for polymer property prediction

Data and results associated with the manuscript "A graph representation of molecular ensembles for polymer property prediction", Chem. Sci. 2022, 13, 10486-10498.

The dataset of computed electron affinities and ionization potentials can be found in the folder datasets/vipea, but is also provided as a csv as part of the Supporting Information. If you'd like to reproduce the calculations, take a look at the files in this GitHub repo, but if you just want the dataset, you can download it from the Electronic SI.

Requirements

To parse all data and recreate the input files you need:

• rdkit >= 2021
• natsort

To reproduce the results:

• chemprop == 1.4.0-polymer
• scikit-learn

Note that the modified version of Chemprop with weighted edges (wD-MPNN, or "Polymer Chemprop") is needed. This can be found in the polymer-chemprop repository. However, to run only the baselines with the standard Chemprop implementation (D-MPNN), chemprop >= 1.3.1 is sufficient.

To reproduce the figures:

• matplotlib
• seaborn

The details of the Python environment that was used for this work can be found in the environment.yml file.

Citation

If you use the dataset or the wD-MPNN version of Chemprop for polymer property prediction please cite

@article{wdmpnn,
         title={A graph representation of molecular ensembles for polymer property prediction}, 
         author={Matteo Aldeghi and Connor W. Coley},
         journal="Chem. Sci.",
         year="2022",
         volume="13",
         issue="35",
         pages="10486-10498",
         publisher="The Royal Society of Chemistry",
         doi="10.1039/D2SC02839E",
         url="http://dx.doi.org/10.1039/D2SC02839E"
}